CN102831579B - Text enhancement method and device, text extraction method and device - Google Patents

Abstract

The embodiment of the present invention discloses a text enhancement method and device, and a text extraction method and device. The text enhancement method includes: acquiring an original image including a line of text; The direct difference degree and indirect difference degree of each neighboring pixel point, the stroke two-dimensional filtering is performed on the original brightness value or/and color value of each original pixel point to obtain the updated brightness value or/and color value of the original image after filtering , the range of the neighborhood set is a square with the original pixel as the center and the side length is w, and the w is smaller than the height of the original image; the updated brightness value or/and color value after the filtering are respectively replaced by the corresponding The original brightness value or/and color value, so as to generate a text-enhanced image corresponding to the original image. Through the embodiments of the present invention, the text contained in the original image can be enhanced, so that the subsequent text extraction for the text-enhanced image is more precise and accurate.

Description

Text enhancement method and device, text extraction method and device

technical field

The present invention generally relates to the technical field of image processing, in particular to a text enhancement method and device, and a text extraction method and device.

Background technique

During the process of playing a video, or when appreciating an image, the video or the image often contains a text description, for example, a description of the time and place where a video occurs, or an explanation of the image. Because these text contents are closely related to videos or images, extracting text from videos or images has become a crucial technology.

The text extraction method of video or image in the prior art can be based on binarization, edge color clustering and detection technology to extract text in the image or video.

However, when the prior art extracts text, there may be too much noise in the video or image, the image or video is blurred, and even if there is a lighting change in a video, the boundary between the text and the background in the image or video will be blurred , or the text content itself is not clear enough, thus affecting the effect of text extraction.

Therefore, how to enhance the text in the original image or video, so as to enhance the text in the image or video, and further optimize the effect of text extraction, has become an urgent problem to be solved in the prior art.

Contents of the invention

In view of this, the embodiment of the present invention provides a text enhancement method and device, and a text extraction method and device, which can enhance the text in the original image including a line of text, so that the text in the original image is more obvious, and can further optimize The effect of text extraction.

According to an aspect of an embodiment of the present invention, a text enhancement method is provided, including: acquiring an original image including a line of text; degree of difference and degree of indirect difference, performing stroke two-dimensional filtering on the original brightness value or/and color value of each original pixel to obtain the filtered updated brightness value or/and color value of the original image, the The range of the neighborhood set is a square centered on the original pixel point and the side length is w, and the w is smaller than the height of the original image; the updated brightness value or/and color value after the filtering are respectively replaced by the corresponding The original brightness value or/and color value to generate a text-enhanced image corresponding to the original image.

According to another aspect of the embodiments of the present invention, there is provided a text enhancement device, including: an acquisition module, configured to acquire an original image including a line of text; a filtering module, configured to filter any original pixel in the original image to The direct difference degree and indirect difference degree of each neighborhood pixel in its neighborhood set, and perform stroke two-dimensional filtering on the original brightness value or/and color value of each original pixel point to obtain the filtered original image Update the brightness value or/and color value, the scope of the neighborhood set is a square centered on the original pixel point and the side length is w, and the w is less than the height of the original image; the replacement module is used to convert the The filtered updated brightness value or/and color value respectively replaces the corresponding original brightness value or/and color value, so as to generate a text-enhanced image corresponding to the original image.

According to yet another aspect of the embodiments of the present invention, a text extraction method is provided, including: acquiring an original image including a line of text; direct difference degree and indirect difference degree, performing two-dimensional stroke filtering on the original luminance value or/and color value of each original pixel, so as to obtain the updated luminance value or/and color value of the original image after filtering; The range of the neighborhood set is a square centered on the original pixel point and the side length is w; the w is smaller than the height of the original image; the updated brightness value or/and color value after the filtering are respectively replaced by the corresponding The original brightness value or/and color value to generate a text-enhanced image corresponding to the original image; extract text in the text-enhanced image.

According to yet another aspect of the embodiments of the present invention, a text extraction device is provided, including: an acquisition module, configured to acquire an original image including a line of text; a filtering module, configured to obtain The direct difference degree and indirect difference degree of each neighborhood pixel in its neighborhood set, and perform stroke two-dimensional filtering on the original brightness value or/and color value of each original pixel point to obtain the filtered original image Update the brightness value or/and color value, the scope of the neighborhood set is a square centered on the original pixel point and the side length is w, and the w is less than the height of the original image; the replacement module is used to convert the The filtered update brightness value or/and color value respectively replaces the corresponding original brightness value or/and color value to generate a text-enhanced image corresponding to the original image; an extraction module is used to extract the text-enhanced image text in .

In addition, according to another aspect of the present invention, a storage medium is also provided. The storage medium includes machine-readable program code. When the program code is executed on the information processing device, the program code causes the information processing device to execute the above text enhancement method and text extraction method according to the present invention.

In addition, according to still another aspect of the present invention, a program product is also provided. The program product includes machine-executable instructions. When the instructions are executed on the information processing device, the instructions cause the information processing device to execute the above text enhancement method and text extraction method according to the present invention.

According to the above-mentioned text enhancement method of the embodiment of the present invention, the text strokes in the pixels in the obtained text enhancement image are enhanced, the consistency of the internal pixels of the strokes is enhanced and the difference between the text and the background is deepened, that is, the original The text included in the image, so that the subsequent text extraction for the text-enhanced image can be more precise and accurate.

According to the above text extraction method of the embodiment of the present invention, text extraction is performed based on the obtained text enhanced image, which not only makes the result of text extraction more precise and accurate, but also reduces the complexity of text extraction, thus improving the text extraction efficiency. s efficiency.

Further aspects of the embodiments of the present invention are given in the description section below, wherein the detailed description serves to fully disclose preferred embodiments of the embodiments of the present invention without imposing limitations thereon.

Description of drawings

The above and other objectives and advantages of the embodiments of the present invention will be further described below in conjunction with specific embodiments and with reference to the accompanying drawings. In the drawings, the same or corresponding technical features or components will be indicated by the same or corresponding reference numerals.

Fig. 1 is a flowchart showing the first text enhancement method provided as an embodiment of the present invention;

Fig. 2 is a flowchart showing step S102 as the first text enhancement method;

Fig. 3 is a flowchart showing a second text enhancement method provided as an embodiment of the present invention;

Fig. 4 is a flow chart showing step S302 as the second text enhancement method;

Fig. 5 is another flow chart showing step S302 as the second text enhancement method;

Fig. 6 is another flow chart showing step S302 in the second text enhancement method;

Fig. 7 is another flow chart showing step S302 as the second text enhancement method;

FIG. 8 is a flow chart showing step S304 as a second text enhancement method;

Fig. 9 is a schematic diagram showing a first text enhancement device provided as an embodiment of the present invention;

FIG. 10 is a schematic diagram showing a filtering module 902 in the first text enhancement device;

Fig. 11 is a schematic diagram showing a second text enhancement device provided as an embodiment of the present invention;

Fig. 12 is a schematic diagram showing the stroke polarity estimation module 1101 in the second text enhancement device;

FIG. 13 is another schematic diagram showing the stroke polarity estimation module 1101 in the second text enhancement device;

Fig. 14 is another schematic diagram showing the stroke polarity estimation module 1101 in the second text enhancement device;

Fig. 15 is another schematic diagram showing the stroke polarity estimation module 1101 in the second text enhancement device;

FIG. 16 is a schematic diagram showing a judgment module 1102 in the second text enhancement device;

Fig. 17 is a flowchart illustrating a text extraction method provided as an embodiment of the present invention;

Fig. 18 is a schematic diagram showing a text extraction device provided as an embodiment of the present invention;

FIG. 19 is a block diagram showing an exemplary structure of a personal computer as an information processing device employed in an embodiment of the present invention.

Detailed ways

Embodiments of the present invention will be described below with reference to the drawings.

Specifically, referring to FIG. 1, the embodiment of the present invention provides a first text enhancement method, which may specifically include:

S101: Acquire an original image including a line of text.

In the embodiment of the present invention, the text enhancement refers to enhancing the text in the original image including a line of text. The enhancement here can be understood as deepening the edge of the text, or highlighting the difference between the text and the background. When the embodiment of the present invention is applied, the stroke appearance (such as brightness or color, etc.) and its shape (such as the text presents stripes) information of the text are considered, in order to achieve the purpose of enhancing the consistency of the internal pixels of the stroke and deepening the difference between the text and the background. Effect.

S102: According to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, the original brightness value or/and color value of each original pixel point is calculated Two-dimensional filtering of strokes to obtain the filtered updated brightness value or/and color value of the original image, the range of the neighborhood set is a square centered on the original pixel and with a side length of w, where w is less than The height of the original image.

The direct difference degree in this step represents the direct appearance difference between any original pixel point and pixels in various fields in its neighborhood set, such as color or brightness difference, and the indirect difference degree represents the difference between the original pixel point and the pixels in each field in its neighborhood set. The gradient modulus of the pixels the point passes through. The original luminance value or/and color value of each original pixel point can be subjected to two-dimensional stroke filtering by using the direct difference degree and the indirect difference degree, so as to obtain the filtered updated luminance value or/and color value of the original image. Among them, the neighborhood set mentioned here can be obtained by selecting a square with the original pixel as the center and w as the side length, where w is less than the image height of the original image, preferably one-eighth of the original image height one.

In practical applications, as shown in FIG. 2, the S102 may specifically include:

S201: Algebraically subtract the original pixel point from the original brightness value or/and color value of each neighboring pixel point to obtain the direct difference degree.

In this embodiment, D ₁ (i, j) is used to represent the direct difference between pixels i and j, then the following formula (1) can be used to calculate the direct difference of brightness values:

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>}{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

Among them, f'(i) represents the mean value of the neighborhood brightness of the target pixel, that is, the direct difference degree is calculated by using the neighborhood pixel mean value instead of the target pixel value; σ(i) represents the local brightness standard deviation around pixel i, which can be normalized effect.

The following formula (2) can also be used to calculate the direct difference between the brightness values of pixels i and j:

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>}{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Among them, f(i) represents the brightness of the neighborhood of the target pixel.

It should be noted that when calculating the direct difference degree of the color values of pixels i and j, formula (3) or formula (4) can be used to calculate:

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>}{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; [</span>{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>}{\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

n in formulas (3) and (4) represents the R channel, G channel and B channel of the color. It should be noted that the above formula for calculating the direct difference degree is just an example, and those skilled in the art can make adaptive changes to the above formula.

S202: Obtain the indirect difference degree according to the gradient modulus from the original pixel point to each neighboring pixel point in its neighborhood set.

Wherein in this embodiment, D ₂ (i, j) is used to represent the indirect difference degree of pixels i and j, then formula (5) can be used to calculate the indirect difference degree of the luminance values of pixels i and j:

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{\mathrm{<span\; class="notranslate">\; dev</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

in, Indicates the gradient modulus at pixel l along the direction i to j. Both b in the formula (5) and a in the formula (1) are pre-set parameters, and the monotonicity of the two is consistent to control the smoothness of the filter.

Of course, in practical applications, the difference between the maximum and minimum gradient values of the brightness of pixels passing through i to j can also be used to replace the maximum gradient value in formula (5), and the calculation method is shown in formula (6):

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{\mathrm{<span\; class="notranslate">\; dev</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; min</span>}}\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{\mathrm{<span\; class="notranslate">\; dev</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; \&sigma;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

Among them, Max represents the maximum value of the gradient value, and min represents the minimum value of the gradient value.

The indirect difference between the color values of pixels i and j can be calculated using formula (7) and formula (8) respectively:

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{{\mathrm{<span\; class="notranslate">\; dev</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

${\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{{\mathrm{<span\; class="notranslate">\; dev</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; min</span>}}\frac{\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\mathrm{<span\; class="notranslate">\; j</span>}}{{\mathrm{<span\; class="notranslate">\; dev</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$

Among them, n represents the R channel, G channel and B channel of the color. It should be noted that the above formula for calculating the indirect difference degree is just an example, and those skilled in the art can make adaptive changes to the above formula.

S203: According to the direct difference degree and the indirect difference degree, calculate the weight value of each neighboring pixel point for the brightness value or/and color value of the original pixel point.

After obtaining the indirect difference degree and direct difference degree, the weight value can be calculated by formula (9):

w(i, j)/w _n (i, j)=exp{-[D ₁ (i, j)+D ₂ (i, j)]} (9)

Among them, D ₁ (i, j) represents the direct difference degree of brightness and/or color value of pixel i and j, and D ₂ (i, j) represents the indirect difference degree of brightness and/or color value of pixel i and j, where , w(i, j) represents the brightness weight value, w _n (i, j) represents the color weight value.

S204: Calculate the updated luminance value of the original pixel by using stroke two-dimensional filtering formula (10).

$\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; )</span>$

Wherein, the N(i) represents the neighborhood set of the pixel i; the w(i, j) represents the weight value of the neighborhood pixel j to the brightness value of the original pixel i; f(j) is the Brightness value of pixel point j in the neighborhood set.

S205: Calculate the updated color value of the original pixel by using stroke two-dimensional filter formula (11).

${\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 11</span>}<span\; class="notranslate">\; )</span>$

Wherein, the w _n (i, j) represents the weight value of the color value of the neighborhood pixel j to the original pixel i on the n channel; f _n (j) is the pixel in the neighborhood set on the n channel The color value of j.

It should be noted that since steps S204 and S205 calculate the updated brightness value and updated color value respectively, in practical applications, any one step can be selected to be executed, or both steps can be executed simultaneously to implement the embodiment of the present invention.

S103: Replace the corresponding original brightness value or/and color value with the filtered updated brightness value or/and color value, respectively, so as to generate a text-enhanced image corresponding to the original image.

After obtaining the updated brightness value or/and color value after filtering, replace the original brightness value or/and color value with the updated brightness value or/and color value respectively, so that the text strokes in the pixels in the original image after replacement are just Enhanced, it can also enhance the consistency of the pixels inside the stroke and deepen the difference between the text and the background.

For the above problems, the embodiment of the present invention also provides a corresponding solution. Specifically, referring to FIG. 3 , the embodiment of the present invention provides another text enhancement method, which may specifically include:

S301: Acquire an original image including a line of text.

S302: Estimate the stroke polarity of the text in the original image, where the polarity represents the relationship between brightness values or/and color values between pixels inside the stroke area and pixels outside the stroke area.

In practical applications, since the stroke enhancement of text is mainly based on filtering technology, that is, the surrounding pixel values outside the stroke are used to enhance the target pixel value inside the stroke, so the noise pixels around the target pixel inside the stroke will have a negative impact on the stroke enhancement effect , this effect is especially noticeable when dealing with thinner strokes or intervals between strokes. In order to prevent such degradation, a stroke polarity estimation scheme is introduced in this embodiment. The stroke polarity estimated in this step may represent the magnitude relationship between the brightness value or/and the color value between the pixels inside the stroke area and the pixels outside the stroke area.

Specifically, in the case where the polarity represents the relationship between the brightness value of the pixel points inside the stroke area and the pixel points outside the stroke area, refer to Figure 4, the step of estimating the stroke polarity of the text in the original image ,include:

S401: Calculate the stroke response intensity using formula (12) in the horizontal direction, vertical direction and two diagonal directions respectively:

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 12</span>}<span\; class="notranslate">\; )</span>$

Wherein, the w is one-eighth of the height of the original image, and the f(i) represents the brightness value of the pixel point i. In this step, four stroke response intensities in the horizontal direction, vertical direction and two diagonal directions can be obtained.

S402: Judging whether the largest stroke response strength among the four calculated stroke response strengths meets the following two conditions: the polarity of [f(i)-f(l)] and [f(i)-f(k)] The same and the stroke response intensity is greater than the preset threshold; if yes, execute step S403; if no, execute step S404.

In a specific application, because the brightness or color value of the pixel inside the text stroke and the background pixel is usually opposite, if [f(i)-f(l)] and [f(i)-f(k) ] have the same polarity, it means that the i pixel point is very likely to be a pixel point inside the stroke. This step is to find out the pixel point that may be inside the stroke from all the pixels in the original image. The polarity here is the same, which means that the brightness value and/or color value of point i is greater than l pixel and k pixel at the same time, or the brightness value and/or color value of i point is less than or equal to l pixel and k pixel at the same time, That is, both [f(i)-f(l)] and [f(i)-f(k)] are greater than zero, or both are less than or equal to zero. The threshold value of stroke response intensity can be adjusted according to actual needs, so the present invention does not limit the selection of the threshold value.

S403: Determine the estimated stroke polarity of the line of text according to the polarity of [f(i)-f(l)] or [f(i)-f(k)].

When the i pixel point satisfies the above two conditions, then determine the estimated stroke polarity in the text according to the polarity of [f(i)-f(l)] or [f(i)-f(k)] as p( i), its value can be arbitrary, it only needs to distinguish the internal pixel points and external pixel points of the stroke. For example, if the brightness value of the strokes in the text is lower than the brightness value of the background pixels, then set p(i) to 0, then correspondingly, p(i) is 1, which means that the brightness value of the strokes in the text is higher than or equal to the background pixels The brightness value of the point.

S404: Select the calculated stroke response strengths in sequence according to the size relationship, and execute the step S402.

When the maximum stroke response intensity does not meet the above two conditions, then select the next largest stroke response intensity to execute the judgment step of step S402, and so on according to the relationship between the stroke response intensity until the two conditions are met or, in the case that none of the four stroke response strengths satisfies the two conditions, the pixel point i is taken as a non-stroke pixel point.

Specifically, in the case where the polarity represents the magnitude relationship between the brightness values of the pixels inside the stroke area and the pixels outside the stroke area, there is another implementation mode. Referring to FIG. 5 , the estimated original image The steps of stroke polarity of Chinese text may specifically include:

S501: Calculate the stroke response intensity of each original pixel in the original image using formula (12) in one direction; the one direction is any one of the horizontal direction, the vertical direction and two diagonal directions.

In this step, the stroke response intensity is firstly calculated in any one of the horizontal direction, the vertical direction and the two diagonal directions.

S502: Determine whether the stroke response intensity satisfies the following two conditions at the same time: [f(i)-f(l)] and [f(i)-f(k)] have the same polarity and the stroke response intensity is greater than The preset threshold, if yes, execute step S503.

Then judge whether the calculated stroke response intensity satisfies the above two conditions at the same time, if not, then do not perform any processing on the calculated stroke response intensity, and then calculate the original image in other directions that have not been calculated The stroke response strength of each original pixel in .

S503: Determine the initial polarity of the original pixel point i according to the polarity of [f(i)-f(l)] or [f(i)-f(k)].

If the calculated stroke response intensity has met the above two conditions, then determine the original pixel point i according to the polarity of [f(i)-f(l)] or [f(i)-f(k)] initial polarity. The initial polarity determined here should be the same as [f(i)-f(l)] or [f(i)-f(k)].

S504: Determine whether all stroke response strengths in the four directions have been calculated, if yes, perform step S505; if not, perform step S501.

Then judge whether all the stroke response strengths in the four directions of horizontal direction, vertical direction and two diagonal directions have been calculated. Execute the step S501.

S505: Determine the initial polarity corresponding to the largest stroke response strength in the four directions as the estimated stroke polarity of the line of text.

If the stroke response strengths in the four directions are all calculated, among the stroke response strengths satisfying the above two conditions, select the initial polarity corresponding to the largest stroke response strength to determine the estimated stroke polarity of the text, namely is that the estimated stroke polarity of the text in the original image is the same as the initial polarity corresponding to the maximum stroke response strength.

Specifically, in the case where the polarity represents the color value size relationship between the pixels inside the stroke area and the pixels outside the stroke area, the step of estimating the stroke polarity of the text in the original image is shown in FIG. 6 , which can include:

S601: Calculate the stroke response intensity of each channel in the horizontal direction, vertical direction and two diagonal directions respectively using the following formulas:

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 13</span>}<span\; class="notranslate">\; )</span>$

Wherein, the w is one-eighth of the height of the original image, and the f _n (i) represents the color value of the pixel point i on the channel n. For example, if the n channels are R channel, G channel and B channel respectively, then the stroke response intensity is the sum of the stroke response intensities on the R channel, G channel and B channel.

S602: Judging whether the largest stroke response strength among the four stroke response strengths of each channel obtained by calculation satisfies the following two conditions: [f _n (i)-f _n (l)] and [f _n (i) on channel n )-f _n (k)] are consistent in polarity and the stroke response intensity is greater than a preset threshold, if yes, execute step S603; if no, execute step S604.

In this step, the condition that the polarities of [f _n (i)-f _n (l)] and [f _n (i)-f _n (k)] must be consistent must be met on any channel.

S603: Determine the estimated stroke polarity of the line of text according to the polarity of [f _n (i)-f _n (l)] or [f _n (i)-f _n (k)].

S604: Select the calculated stroke response strengths sequentially according to the size relationship, and execute the step S602.

When the largest stroke response intensity does not meet the above two conditions, this step needs to select the next largest, third and fourth largest stroke response intensity according to the size relationship to perform step S602 until a certain stroke response intensity satisfies the described Two conditions, or, in the case that the four stroke response intensities do not satisfy the two conditions, the pixel i is regarded as a non-stroke pixel. For example, when the next largest stroke response intensity meets the above two conditions, the stroke polarity estimation process is stopped.

In practical applications, in the case where the polarity represents the color value size relationship between the pixels inside the stroke area and the pixels outside the stroke area, there is another scenario, as shown in Figure 7, the estimated original Steps for stroke polarity of text in an image, including:

S701: Use formula (13) to calculate the stroke response strength of each channel of each original pixel point in the original image in one direction, the one direction being any one of the horizontal direction, the vertical direction and the two diagonal directions direction.

S702: Determine whether the stroke response intensity of each channel in the one direction satisfies the following two conditions at the same time: [f _n (i)-f _n (l)] and [f _n (i)-f _n ( The polarities of k)] are consistent and the stroke response intensity is greater than the preset threshold, if yes, execute step S703.

This step directly judges whether a stroke response intensity calculated in step S701 satisfies the two conditions, and if not, no initial polarity is set in this direction.

S703: Determine the initial polarity of the original pixel i according to the polarity of [f _n (i)-f _n (l)] or [f _n (i)-f _n (k)].

If a stroke response intensity calculated in step S701 satisfies the two conditions, then the initial polarity of the original pixel i in this direction is set to be consistent with [f _n (i)-f _n (l)] or [f _n (i)-f _n (k)] have the same polarity.

S704: Determine whether all stroke response strengths in the four directions have been calculated, if yes, execute step S705; if not, execute step S701

S705: Determine the initial polarity corresponding to the maximum stroke response strength in the four directions as the estimated stroke polarity of the line of text.

If the stroke response strengths in the four directions are all calculated, select the initial polarity corresponding to the largest stroke response strength from the stroke response strengths in the four directions satisfying the above two conditions, as the estimated stroke of the text polarity. If the stroke response strengths in the four directions have not all been calculated, then select any direction that has not been calculated to execute the step S701.

S303: According to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, the original brightness value or/and color value of each original pixel point is calculated Stroke two-dimensional filtering, to obtain the filtered updated brightness value or/and color value of the original image, the range of the neighborhood set is a square with the original pixel as the center and a side length of w, and the w is less than The height of the original image.

For a detailed introduction of this step, reference may be made to relevant content in Embodiment 1, and details are not repeated here. It should be noted that this step can be executed simultaneously with step S302, or step S303 can be executed first, and then step S302 can be executed.

S304: For each pixel of the original image, determine whether the filtered updated brightness value or/and color value matches the stroke polarity, if yes, execute step S305, if not, do not replace.

After the stroke polarity is calculated and stroke filtering is performed, it can be judged in turn whether the filtered updated brightness value or/and color value of each pixel of the original image matches the stroke polarity. If the updated brightness value or/and color value does not match the stroke polarity, the original brightness value or/and color value of the pixel point will not be replaced, and continue to judge the filtered updated brightness value or // and whether the color value matches said stroke polarity.

Referring to FIG. 8, the step S304 may specifically include:

S801: Obtain a first magnitude relationship between the filtered updated brightness value or/and color value and the original brightness value or/and color value.

First obtain the first size relationship between the filtered updated brightness value or/and color value and the original brightness value or/and color value, the first size relationship here can be, for example, the brightness of the updated brightness value is brighter than the original brightness value, Then the first magnitude relationship is that the updated brightness value is greater than the original brightness value, or, for example, the updated color value is greater than the original color value, then the first magnitude relationship is that the updated color value is greater than the original color value. The first size relationship can be obtained by algebraically subtracting the updated brightness value or/and color value from the original brightness value or/and color value.

S802: Determine whether the first size relationship matches the second size relationship represented by the stroke polarity.

The stroke polarity represents the relationship between the brightness value or/and color value of the pixel inside the stroke and the brightness value or/and color value of the pixel outside the stroke, therefore, when the updated color value is larger than the original color value , and the second size relationship represented by the stroke polarity is also when the updated color value is larger than the original color value, or when the updated color value is smaller than the original color value, and the second size relationship represented by the stroke polarity is also the updated color If the value is smaller than the original color value, it is considered that the first size relationship matches the second size relationship represented by the stroke polarity; otherwise, it is considered that the first size relationship does not match the second size relationship.

S305: Replace the corresponding original brightness value or/and color value with the filtered updated brightness value or/and color value, respectively, so as to generate a text-enhanced image corresponding to the original image.

In the case that the first size relationship matches the second size relationship, the corresponding original brightness value or/and color value can be replaced by the updated brightness value or/and color value after filtering, so as to obtain the The text augmented image corresponding to the original image. In this embodiment, through the steps of stroke polarity estimation, text enhancement can be performed when the stroke polarity matches the updated brightness value or color value after filtering. Compared with embodiment 1, the enhanced text The effect is more prominent, which is more conducive to the improvement of the accuracy of subsequent text extraction.

Corresponding to the first text enhancement method provided in the embodiment of the present invention, the embodiment of the present invention also provides a text enhancement device, see FIG. 9, the device may include:

An acquisition module 901, configured to acquire an original image including a line of text.

The filtering module 902 is configured to, according to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, calculate the original brightness value or/or of each original pixel point Perform two-dimensional filtering of strokes and color values to obtain filtered update brightness values or/and color values of the original image, the range of the neighborhood set is a square centered on the original pixel point and a side length of w, The w is smaller than the height of the original image.

Wherein, referring to FIG. 10, the filtering module 902 may specifically include:

The first acquisition sub-module 1001 is used to algebraically subtract the original pixel point from the original brightness value or/and color value of each neighboring pixel point to obtain the direct difference degree;

The second acquisition sub-module 1002 is used to acquire the indirect difference degree according to the gradient modulus from the original pixel point to each neighboring pixel point in its neighborhood set;

The weight calculation sub-module 1003 is used to calculate the weight value of each neighboring pixel point for the brightness value or/and color value of the original pixel point according to the direct difference degree and the indirect difference degree;

The updated luminance value calculation sub-module 1004 is used to calculate the updated luminance value of the original pixel by using stroke two-dimensional filtering formula (10).

The updated color value calculation sub-module 1005 is used to calculate the updated color value of the original pixel by using stroke two-dimensional filtering formula (11).

The replacement module 903 is configured to replace the corresponding original brightness value or/and color value with the updated brightness value or/and color value after filtering, so as to generate a text-enhanced image corresponding to the original image.

Through the above-mentioned text enhancement device provided by the embodiment of the present invention, after obtaining the updated brightness value or/and color value after filtering, replace the original brightness value or/and color value with the updated brightness value or/and color value respectively, thus replacing Afterwards, the text strokes in the pixels in the original image are enhanced, which can also enhance the consistency of the pixels inside the strokes and deepen the difference between the text and the background, thereby providing better text enhancement images for subsequent text extraction, and can also improve subsequent text extraction. Accuracy and precision of text extraction.

Corresponding to the second text enhancement method provided by the embodiment of the present invention, the embodiment of the present invention also provides a text enhancement device, see FIG. 11, the device may include:

An acquisition module 901, configured to acquire an original image including a line of text.

The stroke polarity estimation module 1101 is configured to estimate the stroke polarity of the text in the original image, and the polarity represents the brightness value or/and the size of the color value between the pixels inside the stroke area and the pixels outside the stroke area relation.

The stroke polarity estimation module 1101 can be configured differently in different application scenarios. In one scenario, in the case where the polarity represents the relationship between the brightness values of the pixels inside the stroke area and the pixels outside the stroke area, as shown in FIG. 12 , the stroke polarity estimation module 1101 may include :

The first calculation sub-module 1201 is used to calculate the stroke response intensity using formula (12) in the horizontal direction, vertical direction and two diagonal directions respectively.

The first judging sub-module 1202 is used to judge whether the maximum stroke response strength among the four calculated stroke response strengths satisfies the following two conditions: [f(i)-f(l)] and [f(i)-f (k)] have the same polarity and the stroke response intensity is greater than a preset threshold.

The first determination sub-module 1203 is used to, in the case that the result of the first determination sub-module is yes, according to [f(i)-f(l)] or [f(i)-f(k)] Polarity determines the estimated stroke polarity of the line of text.

The first triggering submodule 1204 is used to select the stroke response strengths obtained according to the size relationship in sequence and trigger the first judging submodule if the result of the first judging submodule is negative, until a certain One stroke response intensity satisfies the two conditions, or, when none of the four stroke response intensities meets the two conditions, the pixel i is regarded as a non-stroke pixel.

In another scenario, in the case where the polarity represents the relationship between brightness values between the pixels inside the stroke area and the pixels outside the stroke area, as shown in FIG. 13 , the stroke polarity estimation module 1101 includes:

The second calculation sub-module 1301 is used to calculate the stroke response strength of each original pixel in the original image using formula (12) in one direction, the one direction being the horizontal direction, the vertical direction and two diagonal lines any of the directions.

The second judging submodule 1302 is used to judge whether the stroke response intensity satisfies the following two conditions at the same time: the polarity of [f(i)-f(l)] is the same as that of [f(i)-f(k)]. And the stroke response intensity is greater than a preset threshold.

The second determination sub-module 1303 is used for determining the polarity according to [f(i)-f(l)] or [f(i)-f(k)] when the second determination sub-module is yes Determine the initial polarity of the original pixel point i.

The third judging sub-module 1304 is used for judging whether all the stroke response strengths in the four directions have been calculated.

The third determination sub-module 1305 is used to determine the initial polarity corresponding to the maximum stroke response strength in the four directions as the estimated stroke of the line of text when the result of the third determination sub-module is yes polarity.

The second triggering submodule 1306 is configured to trigger the second calculation submodule when the result of the third judging submodule is negative.

In another scenario, in the case where the polarity represents the color value magnitude relationship between the pixels inside the stroke area and the pixels outside the stroke area, as shown in FIG. 14 , the stroke polarity estimation module 1101 can specifically include:

The third calculation sub-module 1401 is used to calculate the stroke response intensity using formula (13) in the horizontal direction, vertical direction and two diagonal directions respectively.

The fourth judging sub-module 1402 is used to judge whether the largest stroke response strength among the calculated stroke response strengths meets the following two conditions: [f _n (i)-f _n (l)] and [f _n The polarities of (i)-f _n (k)] are consistent and the stroke response intensity is greater than a preset threshold.

The fourth determination submodule 1403 is configured to, when the result of the fourth determination submodule is yes, according to [f _n (i)-f _n (l)] or [f _n (i)-f _n ( k)] determines the estimated stroke polarity of the line of text.

The third triggering submodule 1404 is configured to select the stroke response strengths obtained according to the size relationship in sequence and trigger the fourth judging submodule in the case that the result of the fourth judging submodule is negative, until a certain One stroke response intensity satisfies the two conditions, or, when none of the four stroke response intensities meets the two conditions, the pixel i is regarded as a non-stroke pixel.

In another scenario, in the case where the polarity represents the color value magnitude relationship between the pixels inside the stroke area and the pixels outside the stroke area, as shown in FIG. 15 , the stroke polarity estimation module 1101 can specifically include:

The fourth calculation sub-module 1501 is used to calculate the stroke response strength of each original pixel in the original image using formula (13) in one direction, the one direction being the horizontal direction, the vertical direction and two diagonal lines any of the directions.

The fifth judging sub-module 1502 is used to judge whether the stroke response intensity satisfies the following two conditions at the same time: [f _n (i)-f _n (l)] and [f _n (i)-f _n on the channel n (k)] are consistent in polarity and the stroke response intensity is greater than a preset threshold.

The fifth determination submodule 1503 is configured to, in the case that the result of the fifth determination submodule is yes, according to [f _n (i)-f _n (i)] or [f _n (i)-f _n ( k)] determine the initial polarity of the original pixel point i.

The sixth judging sub-module 1504 is used to judge whether all the stroke response strengths in the four directions have been calculated.

The sixth determination sub-module 1505 is configured to determine the initial polarity corresponding to the maximum stroke response strength in the four directions as the estimated stroke of the line of text when the result of the sixth determination sub-module is Yes polarity.

The fourth triggering submodule 1506 is configured to trigger the fourth calculation submodule when the result of the sixth judging submodule is negative.

The filtering module 902 is configured to, according to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, calculate the original brightness value or/or of each original pixel point Perform two-dimensional filtering of strokes and color values to obtain filtered update brightness values or/and color values of the original image, the range of the neighborhood set is a square centered on the original pixel point and a side length of w, The w is smaller than the height of the original image.

The judging module 1102 is used to judge whether the filtered updated brightness value or/and color value matches the stroke polarity, and if so, trigger the replacement module 903 .

Referring to FIG. 16, in practical applications, the judging module 1102 may specifically include:

The third obtaining sub-module 1601 is configured to obtain a first magnitude relationship between the filtered updated brightness value or/and color value and the original brightness value or/and color value.

The seventh judging sub-module 1602 is configured to judge whether the first size relationship matches the second size relationship represented by the stroke polarity.

The replacement module 903 is configured to replace the corresponding original brightness value or/and color value with the updated brightness value or/and color value after filtering, so as to generate a text-enhanced image corresponding to the original image.

Through the above-mentioned device provided by the embodiment of the present invention, the stroke polarity estimation method can be further used to verify the updated color value or updated brightness value after filtering, and the stroke polarity and the updated color value and/or updated brightness value after filtering If the values match, the filtered updated brightness value or/and color value is then replaced with the corresponding original brightness value or/and color value, so that the text-enhanced image obtained in this way is more effective and accurate.

In addition, as shown in FIG. 17, after the enhancement of this text, the embodiment of the present invention also provides a text extraction method, which may include:

S1701: Acquire an original image including a line of text.

S1702: According to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, perform an original brightness value or/and color value of each original pixel point Two-dimensional filtering of strokes to obtain the filtered updated brightness value or/and color value of the original image; the range of the neighborhood set is a square with the original pixel as the center and a side length of w; the w is less than The height of the original image.

S1703: Replace the filtered updated brightness value or/and color value with the corresponding original brightness value or/and color value, respectively, to generate a text-enhanced image corresponding to the original image;

S1704: Extract text in the text-enhanced image.

Using the above text extraction method, it can be performed based on text enhancement images, so that the extracted text will be more accurate and precise. At the same time, because the text has been enhanced during extraction, it can also reduce the complexity of text extraction and improve the efficiency of text extraction. .

Corresponding to the above text extraction method, as shown in FIG. 18, an embodiment of the present invention also provides a text extraction device, which may include:

An acquisition module 901, configured to acquire an original image including a line of text.

The filtering module 902 is configured to, according to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, calculate the original brightness value or/or of each original pixel point Perform two-dimensional filtering of strokes and color values to obtain filtered update brightness values or/and color values of the original image, the range of the neighborhood set is a square centered on the original pixel point and a side length of w, The w is smaller than the height of the original image.

The replacement module 903 is configured to replace the corresponding original brightness value or/and color value with the updated brightness value or/and color value after filtering, so as to generate a text-enhanced image corresponding to the original image.

An extraction module 1801, configured to extract text in the text-enhanced image.

Using the above text extraction device can be performed based on text enhancement images, so that the extracted text will be more accurate and precise. At the same time, because the text has been enhanced during extraction, it can also reduce the complexity of text extraction and improve the efficiency of text extraction. .

In addition, it should also be noted that the series of processes and devices described above may also be implemented by software and/or firmware. In the case of realization by software and/or firmware, the program constituting the software is installed from a storage medium or a network to a computer having a dedicated hardware configuration, such as a general-purpose personal computer 1900 shown in FIG. , can perform various functions and so on.

In FIG. 19 , a central processing unit (CPU) 1901 executes various processes according to programs stored in a read only memory (ROM) 1902 or programs loaded from a storage section 1908 to a random access memory (RAM) 1903 . In the RAM 1903, data required when the CPU 1901 executes various processing and the like is also stored as necessary.

The CPU 1901, ROM 1902, and RAM 1903 are connected to each other via a bus 1904. The input/output interface 1905 is also connected to the bus 1904 .

The following components are connected to the input/output interface 1905: an input section 1906 including a keyboard, a mouse, etc.; an output section 1907 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; The storage section 1908 includes a hard disk and the like; and the communication section 1909 includes a network interface card such as a LAN card, a modem, and the like. The communication section 1909 performs communication processing via a network such as the Internet.

A driver 1910 is also connected to the input/output interface 1905 as needed. A removable medium 1911 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 1910 as necessary, so that a computer program read therefrom is installed into the storage section 1908 as necessary.

In the case of realizing the above-described series of processing by software, the programs constituting the software are installed from a network such as the Internet or a storage medium such as the removable medium 1911 .

Those skilled in the art should understand that such a storage medium is not limited to the removable medium 1911 shown in FIG. 19 in which the program is stored and distributed separately from the device to provide the program to the user. Examples of the removable media 1911 include magnetic disks (including floppy disks (registered trademark)), optical disks (including compact disk read only memory (CD-ROM) and digital versatile disks (DVD)), magneto-optical disks (including )) and semiconductor memory. Alternatively, the storage medium may be a ROM 1902, a hard disk contained in the storage section 1908, or the like, in which the programs are stored and distributed to users together with devices containing them.

It should also be pointed out that the steps for executing the above series of processes can naturally be executed in chronological order according to the illustrated order, but it does not need to be executed in chronological order. Certain steps may be performed in parallel or independently of each other.

Although the present invention and its advantages have been described in detail, it should be understood that various changes, substitutions and alterations can be made hereto without departing from the spirit and scope of the invention as defined by the appended claims. Moreover, the terms "comprising", "comprising" or any other variation thereof in the embodiments of the present invention are intended to cover a non-exclusive inclusion such that a process, method, article or device comprising a series of elements includes not only those elements, but also Including other elements not expressly listed, or also including elements inherent in such process, method, article or apparatus. Without further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article or apparatus comprising said element.

Regarding the implementation manner comprising the above embodiments, the following additional notes are also disclosed:

Additional Note 1. A text enhancement method, comprising:

Get the original image including a line of text;

According to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, the stroke two is performed on the original brightness value or/and color value of each original pixel point Dimensional filtering to obtain the filtered updated brightness value or/and color value of the original image, the range of the neighborhood set is a square with the original pixel as the center and a side length of w, and the w is smaller than the the height of the original image;

replacing the corresponding original brightness value or/and color value with the updated brightness value or/and color value after filtering, so as to generate a text-enhanced image corresponding to the original image.

2. The method according to Supplement 1, wherein the step of performing two-dimensional stroke filtering on the original brightness value or/and color value of each original pixel point includes:

Perform algebraic subtraction of the original pixel point and the original brightness value or/and color value of each neighboring pixel point to obtain the direct difference degree;

Obtain the indirect difference degree according to the gradient modulus from the original pixel point to each neighboring pixel point in its neighborhood set;

Calculate the weight value of each neighboring pixel point for the brightness value or/and color value of the original pixel point according to the direct difference degree and the indirect difference degree;

The updated brightness value of the original pixel is calculated by using the following stroke two-dimensional filtering formula;

$\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$

Wherein, the N(i) represents the neighborhood set of the pixel i; the w(i, j) represents the weight value of the neighborhood pixel j to the brightness value of the original pixel i; f(j) is the The brightness value of pixel point j in the neighborhood set;

and / or,

The updated color value of the original pixel is calculated by using the following stroke two-dimensional filtering formula;

${\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$

Wherein, the w _n (i, j) represents the weight value of the color value of the neighborhood pixel j to the original pixel i on the n channel; f _n (j) is the pixel in the neighborhood set on the n channel The color value of j.

3. The method according to Note 1, wherein, after acquiring the original image including a line of text, further comprising:

Estimating the stroke polarity of the text in the original image, the polarity representing the magnitude relationship between the brightness value or/and color value between the pixels inside the stroke area and the pixels outside the stroke area;

And after obtaining the filtered updated brightness value or/and color value of the original image, it also includes:

Judging whether the updated brightness value or/and color value after filtering matches the polarity of the stroke, if so, performing the step of replacing the original brightness value or/and color value.

4. The method according to Note 3, wherein, in the case where the polarity represents the relationship between the brightness value of the pixels inside the stroke area and the pixels outside the stroke area, the estimated strokes of the text in the original image Polarity steps, including:

The following formulas are used to calculate the stroke response strength in the horizontal direction, vertical direction and two diagonal directions respectively:

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$

Wherein, the w is one-eighth of the height of the original image, and the f(i) represents the brightness value of the pixel i;

Judging whether the maximum stroke response intensity among the four calculated stroke response intensity meets the following two conditions: [f(i)-f(l)] has the same polarity as [f(i)-f(k)] and The stroke response intensity is greater than a preset threshold, if so, determine the estimated stroke polarity of the line of text according to the polarity of [f(i)-f(l)] or [f(i)-f(k)] If not, then select the calculated stroke response strength according to the magnitude relationship to execute the judgment step until a certain stroke response strength satisfies the two conditions, or, when none of the four stroke response strengths When the two conditions are satisfied, the pixel i is regarded as a non-stroke pixel.

5. The method according to supplementary note 3, wherein, in the case where the polarity represents the relationship between the brightness value of the pixels inside the stroke area and the pixels outside the stroke area, the estimated strokes of the text in the original image Polarity steps, including:

In one direction, the following formula is used to calculate the stroke response strength of each original pixel in the original image; the one direction is any direction in the horizontal direction, the vertical direction and the two diagonal directions;

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$

Wherein, the w is one-eighth of the height of the original image, and the f(i) represents the brightness value of the pixel i;

Judging whether the stroke response strength meets the following two conditions at the same time: [f(i)-f(l)] and [f(i)-f(k)] have the same polarity and the stroke response strength is greater than the preset Threshold, if yes, then determine the initial polarity of the original pixel i according to the polarity of [f(i)-f(l)] or [f(i)-f(k)];

Determine whether all stroke response strengths in the four directions have been calculated, and if so, determine the initial polarity corresponding to the maximum stroke response strength in the four directions as the estimated stroke polarity of the line of text; if not, then Repeating the step of calculating the stroke response intensity of each original pixel in the original image.

6. The method according to Note 3, wherein, in the case where the polarity represents the color value magnitude relationship between the pixels inside the stroke area and the pixels outside the stroke area, the estimated strokes of the text in the original image Polarity steps, including:

The following formulas are used to calculate the stroke response strength in the horizontal direction, vertical direction and two diagonal directions respectively:

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$

Wherein, the w is one-eighth of the height of the original image, and the f _n (i) represents the color value of the pixel i on the channel n;

Judging whether the maximum stroke response strength among the four stroke response strengths obtained by calculation satisfies the following two conditions: [f _n (i)-f _n (l)] and [f _n (i)-f _n ( The polarities of k)] are consistent and the stroke response intensity is greater than the preset threshold, if so, according to [f _n (i)-f _n (l)] or [f _n (i)-f _n (k )] to determine the estimated stroke polarity of the line of text; if not, then select the calculated stroke response strength according to the size relationship to execute the judgment step until a certain stroke response strength satisfies the two condition, or, in the case that none of the four stroke response intensities satisfies the two conditions, the pixel i is regarded as a non-stroke pixel.

7. The method according to Note 3, wherein, in the case where the polarity represents the color value magnitude relationship between the pixels inside the stroke area and the pixels outside the stroke area, the estimated strokes of the text in the original image Polarity steps, including:

In one direction, the following formula is used to calculate the stroke response strength of each original pixel in the original image, and the one direction is any direction in the horizontal direction, the vertical direction and the two diagonal directions;

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$

Wherein, the w is one-eighth of the height of the original image, and the f _n (i) represents the color value of the pixel i on the channel n;

Determine whether the stroke response strength meets the following two conditions at the same time: the polarities of [f _n (i)-f _n (l)] and [f _n (i)-f _n (k)] on the channel n are maintained consistent _and the stroke response intensity _is greater than the preset _threshold , if so, _determine the The initial polarity of the original pixel point i;

Determine whether all stroke response strengths in the four directions have been calculated, and if so, determine the initial polarity corresponding to the maximum stroke response strength in the four directions as the estimated stroke polarity of the line of text; if not, then Repeating the step of calculating the stroke response intensity of each original pixel in the original image.

8. The method according to Note 3, wherein the step of judging whether the filtered updated brightness value or/and color value matches the polarity of the stroke includes:

Acquiring a first magnitude relationship between the filtered updated brightness value or/and color value and the original brightness value or/and color value;

It is judged whether the first size relationship matches the second size relationship represented by the stroke polarity.

9. A text enhancement device, comprising:

An acquisition module, configured to acquire an original image including a line of text;

A filtering module, configured to perform an original luminance value or/and The color value is subjected to stroke two-dimensional filtering to obtain the filtered updated brightness value or/and color value of the original image, and the range of the neighborhood set is a square with the original pixel as the center and a side length of w, so Said w is less than the height of said original image;

A replacement module, configured to respectively replace the corresponding original brightness value or/and color value with the filtered updated brightness value or/and color value, so as to generate a text-enhanced image corresponding to the original image.

10. The device according to supplementary note 9, wherein the filtering module includes:

The first acquisition submodule is used to algebraically subtract the original pixel point from the original brightness value or/and color value of each neighboring pixel point to obtain the direct difference degree;

The second acquisition sub-module is used to acquire the indirect difference degree according to the gradient modulus from the original pixel point to each neighborhood pixel point of its neighborhood set;

The weight calculation sub-module is used to calculate the weight value of each neighborhood pixel for the brightness value or/and color value of the original pixel according to the direct difference degree and the indirect difference degree;

An updated brightness value calculation sub-module, used to calculate the updated brightness value of the original pixel by using the following stroke two-dimensional filtering formula;

$\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$

Wherein, the N(i) represents the neighborhood set of the pixel i, the w(i, j) represents the weight value of the neighborhood pixel j to the brightness value of the original pixel i, and f(j) is the The brightness value of pixel point j in the neighborhood set; and/or,

The updated color value calculation submodule is used to calculate the updated color value of the original pixel by using the following stroke two-dimensional filtering formula;

${\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$

Wherein, the w _n (i, j) represents the weight value of the color value of the neighborhood pixel point j to the original pixel point i on the n channel, and f _n (j) is the pixel point in the neighborhood set on the n channel The color value of j.

11. The device according to appendix 9, further comprising:

The stroke polarity estimation module is used to estimate the stroke polarity of the text in the original image, and the polarity represents the magnitude relationship between the brightness value or/and the color value between the pixels inside the stroke area and the pixels outside the stroke area ;as well as

The judging module is used to judge whether the filtered updated brightness value or/and color value matches the polarity of the stroke, and if so, trigger the replacing module.

12. According to the device described in supplementary note 11, in the case where the polarity represents the relationship between the brightness value of the pixels inside the stroke area and the pixels outside the stroke area, the stroke polarity estimation module includes:

The first calculation sub-module is used to calculate the stroke response strength using the following formulas in the horizontal direction, the vertical direction and the two diagonal directions respectively:

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$

Wherein, the w is one-eighth of the height of the original image, and the f(i) represents the brightness value of the pixel i;

The first judging submodule is used to judge whether the maximum stroke response strength among the four stroke response strengths obtained by the calculation satisfies the following two conditions: [f(i)-f(l)] and [f(i)-f( k)] have the same polarity and the stroke response intensity is greater than a preset threshold;

The first determining submodule is used for determining the pole according to [f(i)-f(l)] or [f(i)-f(k)] when the result of the first judging submodule is yes determining an estimated stroke polarity of the line of text;

The first triggering submodule is used to select the stroke response strengths obtained according to the size relationship and trigger the first judging submodule sequentially until a certain The stroke response intensity satisfies the two conditions, or, when none of the four stroke response intensities meets the two conditions, the pixel i is regarded as a non-stroke pixel.

13. According to the device described in Supplementary Note 11, in the case where the polarity represents the relationship between the brightness value of the pixels inside the stroke area and the pixels outside the stroke area, the stroke polarity estimation module includes:

The second calculation sub-module is used to calculate the stroke response strength of each original pixel in the original image using the following formula in one direction, the one direction being the horizontal direction, the vertical direction and the two diagonal directions either direction;

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$

Wherein, the w is one-eighth of the height of the original image, and the f(i) represents the brightness value of the pixel i;

The second judging submodule is used to judge whether the stroke response intensity satisfies the following two conditions at the same time: [f(i)-f(l)] and [f(i)-f(k)] have the same polarity and The stroke response intensity is greater than a preset threshold;

The second determining submodule is used to determine the polarity according to [f(i)-f(l)] or [f(i)-f(k)] when the second judging submodule is yes The initial polarity of the original pixel point i;

The third judging sub-module is used to judge whether the stroke response strengths in the four directions have all been calculated;

The third determining submodule is used to determine the initial polarity corresponding to the maximum stroke response strength in the four directions as the estimated stroke pole of the line of text when the result of the third judging submodule is Yes sex;

The second triggering submodule is configured to trigger the second calculation submodule when the result of the third judging submodule is negative.

14. According to the device described in supplementary note 11, in the case where the polarity represents the color value magnitude relationship between the pixels inside the stroke area and the pixels outside the stroke area, the stroke polarity estimation module includes:

The third calculation sub-module is used to calculate the stroke response strength using the following formulas in the horizontal direction, the vertical direction and the two diagonal directions respectively:

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$

Wherein, the w is one-eighth of the height of the original image, and the f _n (i) represents the color value of the pixel i on the channel n;

The fourth judging submodule is used to judge whether the maximum stroke response strength in the calculated stroke response strength meets the following two conditions: [f _n (i)-f _n (l)] and [f _n ( The polarities of i)-f _n (k)] are consistent and the stroke response intensity is greater than a preset threshold;

The fourth determining submodule is used to, when the result of the fourth judging submodule is yes, according to [f _n (i)-f _n (l)] or [f _n (i)-f _n (k )] determine the estimated stroke polarity of the line of text;

The third triggering submodule is used to select the stroke response strengths obtained according to the size relationship and trigger the fourth judging submodule in sequence, until the result of the fourth judging submodule is negative, until a certain The stroke response intensity satisfies the two conditions, or, when none of the four stroke response intensities meets the two conditions, the pixel i is regarded as a non-stroke pixel.

15. According to the device described in supplementary note 11, in the case where the polarity represents the color value magnitude relationship between the pixels inside the stroke area and the pixels outside the stroke area, the stroke polarity estimation module includes:

The fourth calculation sub-module is used to calculate the stroke response strength of each original pixel in the original image using the following formula in one direction, the one direction being the horizontal direction, the vertical direction and the two diagonal directions either direction;

$\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$

Wherein, the w is one-eighth of the height of the original image, and the f _n (i) represents the color value of the pixel i on the channel n;

The fifth judging submodule is used to judge whether the stroke response intensity satisfies the following two conditions at the same time: [f _n (i)-f _n (l)] and [f _n (i)-f _n ( k)] polarities are consistent and the stroke response intensity is greater than a preset threshold;

The fifth determining submodule is used to, when the result of the fifth judging submodule is yes, according to [f _n (i)-f _n (l)] or [f _n (i)-f _n (k )] to determine the initial polarity of the original pixel i;

The sixth judging sub-module is used to judge whether all the stroke response strengths in the four directions have been calculated;

The sixth determining submodule is used to determine the initial polarity corresponding to the maximum stroke response strength in the four directions as the estimated stroke polarity of the line of text when the result of the judging submodule is yes;

The fourth triggering submodule is configured to trigger the fourth calculating submodule when the result of the sixth judging submodule is negative.

16. The device according to supplementary note 11, the judging module includes:

A third acquiring submodule, configured to acquire a first magnitude relationship between the filtered updated brightness value or/and color value and the original brightness value or/and color value;

A seventh judging submodule, configured to judge whether the first size relationship matches the second size relationship represented by the stroke polarity.

17. A text extraction method, comprising:

Get the original image including a line of text;

According to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, the stroke two is performed on the original brightness value or/and color value of each original pixel point Dimensional filtering to obtain the filtered updated brightness value or/and color value of the original image; the range of the neighborhood set is a square with the original pixel as the center and a side length of w; the w is smaller than the the height of the original image;

Replacing the filtered updated brightness value or/and color value with the corresponding original brightness value or/and color value, respectively, to generate a text-enhanced image corresponding to the original image;

Text in the text-enhanced image is extracted.

18. A text extraction device, comprising:

An acquisition module, configured to acquire an original image including a line of text;

A filtering module, configured to perform an original luminance value or/and The color value is subjected to stroke two-dimensional filtering to obtain the filtered updated brightness value or/and color value of the original image, and the range of the neighborhood set is a square with the original pixel as the center and a side length of w, so Said w is less than the height of said original image;

A replacement module, configured to replace the corresponding original brightness value or/and color value with the updated brightness value or/and color value after filtering, so as to generate a text-enhanced image corresponding to the original image;

An extraction module, configured to extract text in the text-enhanced image.

Claims (6)

1. A text enhancement method, comprising: Get the original image including a line of text; According to the direct difference degree and indirect difference degree between any original pixel point in the original image and each neighboring pixel point in its neighborhood set, the stroke two is performed on the original brightness value or/and color value of each original pixel point Dimensional filtering to obtain the filtered updated brightness value or/and color value of the original image, the range of the neighborhood set is a square with the original pixel as the center and a side length of w, and the w is smaller than the the height of the original image; replacing the corresponding original brightness value or/and color value with the filtered updated brightness value or/and color value, respectively, to generate a text-enhanced image corresponding to the original image, Wherein, after the acquisition of the original image including a line of text, it also includes: Estimating the stroke polarity of the text in the original image, the polarity representing the magnitude relationship between the brightness value or/and color value between the pixels inside the stroke area and the pixels outside the stroke area; And after obtaining the filtered updated brightness value or/and color value of the original image, it also includes: judging whether the filtered updated brightness value or/and color value matches the stroke polarity, and if so, performing the step of replacing the original brightness value or/and color value, Wherein, the step of performing stroke two-dimensional filtering on the original brightness value or/and color value of each original pixel point includes: Perform algebraic subtraction of the original pixel point and the original brightness value or/and color value of each neighboring pixel point to obtain the direct difference degree; Obtain the indirect difference degree according to the gradient modulus from the original pixel point to each neighboring pixel point in its neighborhood set; Calculate the weight value of each neighboring pixel point for the brightness value or/and color value of the original pixel point according to the direct difference degree and the indirect difference degree; The updated brightness value of the original pixel is calculated by using the following stroke two-dimensional filtering formula; $\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$ Wherein, the N(i) represents the neighborhood set of the pixel i; the w(i, j) represents the weight value of the neighborhood pixel j to the brightness value of the original pixel i; f(j) is the The brightness value of pixel point j in the neighborhood set; and / or The updated color value of the original pixel is calculated by using the following stroke two-dimensional filtering formula; ${\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span>{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&Center\; Dot;</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$ Wherein, the w _n (i, j) represents the weight value of the color value of the neighborhood pixel j to the original pixel i on the n channel; f _n (j) is the pixel in the neighborhood set on the n channel The color value of j. 2. The method according to claim 1, wherein, under the condition that the polarity represents the brightness value magnitude relationship between the internal pixels of the stroke area and the external pixels of the stroke area, the stroke of the text in the estimated original image Polarity steps, including: The following formulas are used to calculate the stroke response strength in the horizontal direction, vertical direction and two diagonal directions respectively: $\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$ Wherein, the w is one-eighth of the height of the original image, and the f(i) represents the brightness value of the pixel i; Determine whether the maximum stroke response intensity among the four calculated stroke response intensity meets the following two conditions: [f(i)-f(l)] has the same polarity as [f(i)-f(k)] and The stroke response intensity is greater than a preset threshold, if so, determine the estimated stroke polarity of the line of text according to the polarity of [f(i)-f(l)] or [f(i)-f(k)] If not, then select the calculated stroke response strength according to the size relationship to execute the judgment step until a certain stroke response strength satisfies the two conditions, or, when none of the four stroke response strengths When the two conditions are satisfied, the pixel point i is regarded as a non-stroke pixel point. 3. The method according to claim 1, wherein, under the condition that the polarity represents the color value magnitude relationship between the internal pixels of the stroke area and the external pixels of the stroke area, the stroke of the text in the estimated original image Polarity steps, including: The following formulas are used to calculate the stroke response strength in the horizontal direction, vertical direction and two diagonal directions respectively: $\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$ Wherein, the w is one-eighth of the height of the original image, and the f _n (i) represents the color value of the pixel i on the channel n; Judging whether the maximum stroke response strength among the four stroke response strengths obtained by calculation satisfies the following two conditions: [f _n (i)-f _n (l)] and [f _n (i)-f _n ( The polarities of k)] are consistent and the stroke response intensity is greater than the preset threshold, if so, according to [f _n (i)-f _n (l)] or [f _n (i)-f _n (k )] to determine the estimated stroke polarity of the line of text; if not, then select the calculated stroke response strength according to the size relationship to execute the judgment step until a certain stroke response strength satisfies the two condition, or, in the case that none of the four stroke response intensities satisfies the two conditions, the pixel i is regarded as a non-stroke pixel. 4. A text enhancement device comprising: An acquisition module, configured to acquire an original image including a line of text; A filtering module, configured to perform an original luminance value or/and The color value is subjected to stroke two-dimensional filtering to obtain the filtered updated brightness value or/and color value of the original image, and the range of the neighborhood set is a square with the original pixel as the center and a side length of w, so Said w is less than the height of said original image; A replacement module, configured to replace the corresponding original brightness value or/and color value with the updated brightness value or/and color value after filtering, so as to generate a text-enhanced image corresponding to the original image; The stroke polarity estimation module is used to estimate the stroke polarity of the text in the original image, and the polarity represents the magnitude relationship between the brightness value or/and the color value between the pixels inside the stroke area and the pixels outside the stroke area ;as well as a judging module, configured to judge whether the filtered updated brightness value or/and color value matches the polarity of the stroke, and if so, trigger the replacement module, Wherein, the filter module includes: The first acquisition submodule is used to algebraically subtract the original pixel point from the original brightness value or/and color value of each neighboring pixel point to obtain the direct difference degree; The second acquisition sub-module is used to acquire the indirect difference degree according to the gradient modulus from the original pixel point to each neighborhood pixel point of its neighborhood set; The weight calculation sub-module is used to calculate the weight value of each neighborhood pixel for the brightness value or/and color value of the original pixel according to the direct difference degree and the indirect difference degree; An updated brightness value calculation sub-module, used to calculate the updated brightness value of the original pixel by using the following stroke two-dimensional filtering formula; $\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$ Wherein, the N(i) represents the neighborhood set of the pixel i, the w(i, j) represents the weight value of the neighborhood pixel j to the brightness value of the original pixel i, and f(j) is the The brightness value of pixel point j in the neighborhood set; and/or The updated color value calculation submodule is used to calculate the updated color value of the original pixel by using the following stroke two-dimensional filtering formula; ${\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; [</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; \&CenterDot;</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; )</span>$ Wherein, the w _n (i, j) represents the weight value of the color value of the neighborhood pixel point j to the original pixel point i on the n channel, and f _n (j) is the pixel point in the neighborhood set on the n channel The color value of j. 5. The device according to claim 4, in the case where the polarity represents the brightness value magnitude relationship between the internal pixels of the stroke area and the external pixels of the stroke area, the stroke polarity estimation module includes: The first calculation sub-module is used to calculate the stroke response strength using the following formulas in the horizontal direction, the vertical direction and the two diagonal directions respectively: $\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$ Wherein, the w is one-eighth of the height of the original image, and the f(i) represents the brightness value of the pixel i; The first judging sub-module is used to judge whether the maximum stroke response strength among the four stroke response strengths obtained by the calculation satisfies the following two conditions: [f(i)-f(l)] and [f(i)-f( k)] have the same polarity and the stroke response intensity is greater than a preset threshold; The first determining submodule is used for determining the pole according to [f(i)-f(l)] or [f(i)-f(k)] when the result of the first judging submodule is yes determining an estimated stroke polarity of the line of text; The first triggering submodule is used to select the stroke response strengths obtained according to the size relationship and trigger the first judging submodule sequentially until a certain The stroke response intensity satisfies the two conditions, or, when none of the four stroke response intensities meets the two conditions, the pixel i is regarded as a non-stroke pixel. 6. The device according to claim 4, in the case where the polarity represents the color value magnitude relationship between the internal pixels of the stroke area and the external pixels of the stroke area, the stroke polarity estimation module includes: The second calculation sub-module is used to calculate the stroke response strength using the following formulas in the horizontal direction, the vertical direction and the two diagonal directions respectively: $\mathrm{<span\; class="notranslate">\; Stroke\; Strength</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; \&Element;</span><span\; class="notranslate">\; \{</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; /</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; \}</span>}{\mathrm{<span\; class="notranslate">\; max</span>}}\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; l</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; +</span><span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>$ Wherein, the w is one-eighth of the height of the original image, and the f _n (i) represents the color value of the pixel i on the channel n; The second judging sub-module is used to judge whether the maximum stroke response intensity among the four stroke response intensity calculated by calculation satisfies the following two conditions: [f _n (i)-f _n (l)] and [f on the channel n The polarities of _n (i)-f _n (k)] are consistent and the stroke response intensity is greater than a preset threshold; The second determining submodule is used to, when the result of the second judging submodule is yes, according to [f _n (i)-f _n (l)] or [f _n (i)-f _n (k )] determine the estimated stroke polarity of the line of text; The second triggering submodule is used to select the stroke response strengths obtained according to the size relationship and trigger the second judging submodule sequentially until a certain The stroke response intensity satisfies the two conditions, or, when none of the four stroke response intensities meets the two conditions, the pixel i is regarded as a non-stroke pixel.