CN112614078A - Image noise reduction method and storage medium - Google Patents
Image noise reduction method and storage medium Download PDFInfo
- Publication number
- CN112614078A CN112614078A CN202011618368.6A CN202011618368A CN112614078A CN 112614078 A CN112614078 A CN 112614078A CN 202011618368 A CN202011618368 A CN 202011618368A CN 112614078 A CN112614078 A CN 112614078A
- Authority
- CN
- China
- Prior art keywords
- pixel
- noise
- salt
- points
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an image noise reduction method and a storage medium, wherein the method comprises the following steps: s1: judging whether the input image contains noise pixel points, and if so, further judging the type of each noise pixel point; s2: according to the types of the noise pixel points, respectively adopting a corresponding noise reduction method to sequentially process the noise pixel points; s3: and outputting the image subjected to noise reduction processing. The corresponding noise reduction method is adopted to process different types of noise pixel points in the input image, so that the detail information of the image can be kept as much as possible, the fuzzy degree of the processed image is greatly reduced, the quality of the image after noise reduction is improved, and the method has the advantages of low algorithm complexity and high applicability.
Description
Technical Field
The present invention relates to the field of image processing, and in particular, to an image denoising method and a storage medium.
Background
With the development of the integrated circuit manufacturing technology and the continuous improvement of the integrated circuit design level, the CMOS image sensor manufactured based on the CMOS integrated circuit process technology is gradually in an advantageous position in the competition of the CCD due to the characteristics of high integration level, low power consumption, small volume, simple process, low cost, short development period, and the like. The CMOS image sensor is widely applied to various photographic products and has wide market application prospect. After the image information in the image sensor is processed by the analog circuit, the image quality may be degraded due to the interference of the imaging device and the external environment noise. In particular, vertical streaks and various light or dark spots, which are gaussian noise and salt and pepper noise, respectively, are generated. The presence of noise has the following effect: (1) serious noise can deform the image and lose the essential data characteristics of the image; (2) the noise may reduce the quality and accuracy of the image data, and may affect the subsequent processing of the image.
In the processing process of the CMOS image sensor, light signals need to be sampled and converted into analog electrical signals through photosensitive pixels, and then the analog electrical signals are output finally through an amplifier and an analog-to-digital conversion unit (ADC). In a series of processes, various noises are introduced to cause the degradation of image quality, wherein Gaussian noise and salt and pepper noise are the most common noises.
In conclusion, the existence of noise has a direct influence on subsequent image segmentation, feature extraction, image recognition and other higher-level processing. An effective method of suppressing noise is therefore of crucial importance for the processing of images. At present, the method for reducing noise of an image is mainly used for processing a spatial domain of the image and mainly comprises intra-frame noise reduction and inter-frame noise reduction.
Intra-frame noise reduction is based on two-dimensional spatial separation of images into linear and non-linear techniques. The mean filtering belongs to a common linear processing technology, a processing object is mainly Gaussian noise, the mean value of pixels in the field is used as the current gray value, noise can be well suppressed under the condition that a neighborhood space is small, but the blurring degree of an image becomes serious along with the enlargement of the neighborhood space. The median filtering is a nonlinear filtering method, has obvious effect on filtering salt-pepper noise, takes the intermediate value of pixels in the neighborhood as the current gray value, has good processing effect under the condition of low noise density, and loses some edge details of the processed image when the filtering window range is large.
At present, the noise reduction methods of images are various, and there are image transform domain processing methods based on Fourier transform or wavelet transform, and also methods based on bilateral filter histogram matching. However, these methods generally have high algorithm complexity, much time consumption, low application range, and less types of processed noise, and are not suitable for real-time processing of images in a camera. In the related image processing, various kinds of noise are rarely processed by corresponding methods respectively, but only one or two main noises are generally processed, so that although the influence of the main noises is reduced, the influence of other noises still exists, and the existence of the noises still influences the image quality all the time. Especially for the mixed noise composed of gaussian noise and salt and pepper noise, neither mean filtering nor median filtering can achieve the expected effect, so it is necessary to improve the two filtering methods to retain the image details as much as possible and improve the image quality.
Disclosure of Invention
Therefore, the invention provides an image denoising method and a storage medium, which are used for solving the problems that the existing image denoising method cannot achieve the expected effect when processing mixed noise and influences the image quality.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention.
To achieve these objects and other advantages and in accordance with the purpose of the invention, as embodied and broadly described herein, a first aspect of the present invention provides an image noise reduction method characterized by comprising the steps of:
s1: judging whether the input image contains noise pixel points, and if so, further judging the type of each noise pixel point;
s2: according to the types of the noise pixel points, respectively adopting a corresponding noise reduction method to sequentially process the noise pixel points;
s3: and outputting the image subjected to noise reduction processing.
Further, the types of the noise pixel points comprise salt and pepper noise points and/or Gaussian noise points.
Step S1 includes: and judging whether the input image contains salt and pepper noise points and/or Gaussian noise points.
Further, the judging whether the input image contains salt and pepper noise points comprises the following steps:
calculating the average value of all pixel points in a filter subblock with a preset size on an input image;
sequentially judging whether the difference between the pixel value of each pixel point in the filtering sub-block and the average value of all the pixel points in the filtering sub-block is greater than a first threshold value or not, and if so, judging that the pixel point in the filtering sub-block is a salt and pepper noise point; otherwise, the pixel is judged to be a salt and pepper noise point
Further, if it is determined that the filter sub-block contains a salt and pepper noise point, step S2 includes:
s21: and according to the difference of the quantity of the salt and pepper noise points in the filter subblocks, performing weighted median filtering processing on each salt and pepper noise point, and taking a pixel value obtained after weighted median filtering processing as an output value of a central pixel point corresponding to the current filter subblock.
Further, "performing weighted median filtering processing on each salt and pepper noise point according to the difference of the number of salt and pepper noise points in the filter sub-block" includes:
when the number of the non-salt-and-pepper noise points in the filter subblock is more than two, sorting according to the pixel value of each non-salt-and-pepper noise point, performing weighted operation on each sorted non-salt-and-pepper noise point according to a set weight value, and taking an operation result as an output value of a central pixel point corresponding to the current filter subblock; the weight value of the non-salt and pepper noise point is in direct proportion to the pixel value of the non-salt and pepper noise point.
Further, the size of the filter subblock is 3X3 subblocks, the 3X3 subblock includes 9 pixel points in 3 rows and 3 columns, the pixel values of the pixel points in the first row are recorded as C1, C2 and C3 from left to right, the pixel values of the pixel points in the second row are recorded as C4, C5 and C6 from left to right, the pixel values of the pixel points in the third row are recorded as C7, C8 and C9 from left to right, and the C5 is a central pixel point;
step S21 includes:
when all the pixel points in the 3X3 sub-block are judged to be pepper-salt noise points, the calculation formula of the output value q of the central pixel point is as follows: q ═ 5 (C2+ C4+ C5+ C6+ C8);
when all pixel points in the 3X3 sub-block are judged not to be salt and pepper noise points, taking the pixel value of the current central pixel point as the output value q of the central pixel point;
when the number of the non-salt-and-pepper noise points in the 3X3 sub-block is judged to be 1, taking the pixel value of the non-salt-and-pepper noise points as the output value q of the central pixel point;
when the number of the non-salt-and-pepper noise points in the 3X3 sub-block is determined to be 2, the calculation formula of the output value q of the central pixel point is as follows: q ═ Cb × b + Ca × a, Ca < Cb; ca and Cb are the pixel values of two non-salt-pepper noise points respectively; a is the weight value corresponding to the Ca, and b is the weight value corresponding to the Cb;
when the number of the non-salt-and-pepper noise points in the 3X3 sub-block is judged to be more than 3, the calculation formula of the output value q of the central pixel point is as follows: q ═ C1 × n1+C2*n2+…Cn*nnWhere C1-Cn are the pixel values of the non-salt-pepper noise points, n1-nnAnd C1-Cn are weight values corresponding to the non-salt-pepper noise points.
Further, the method comprises:
after all pixel points in the filtering subblock are processed by the salt and pepper noise point, whether the filtering subblock contains a Gaussian noise point or not is judged, and the method specifically comprises the following steps: and calculating whether the absolute value of the pixel gradient in the filter subblock is smaller than a second threshold value, and if so, judging that the filter subblock contains a Gaussian noise point.
Further, the method comprises:
when the filter subblocks contain Gaussian noise points, performing weighted mean filtering processing on the output values of the central pixel points obtained by salt and pepper noise processing to obtain the final output values of the central pixel points of the current filter subblocks;
the weighted mean filtering process includes: configuring a first weight value for a first pixel array in the filtering sub-block and a second weight value for a second pixel array in the filtering sub-block;
the final output value of the central pixel point of the current filter subblock is the pixel mean value of all pixel points of the first pixel array, the first weight value, the pixel mean value of all pixel points of the second pixel array, the second weight value and the output value of the central pixel point obtained by salt and pepper noise processing;
the first pixel array is a pixel point which is transversely adjacent or longitudinally adjacent to a central pixel point in the current filter subblock, and the second pixel array is a pixel point except the first pixel array and the central pixel point in the current filter subblock.
Further, the method comprises:
and moving the positions of the filtering sub-blocks in the input image according to a preset moving sequence, judging the types of noise pixel points in the new filtering sub-blocks, and performing noise reduction processing by adopting a corresponding noise reduction method until all pixel points on the input image are completely traversed, so as to obtain an image subjected to noise reduction processing and output the image.
The second aspect of the present invention also provides a storage medium storing a computer program which, when executed by a processor, performs the method according to the first aspect of the present invention.
The invention discloses an image noise reduction method and a storage medium, wherein the method comprises the following steps: s1: judging whether the input image contains noise pixel points, and if so, further judging the type of each noise pixel point; s2: according to the types of the noise pixel points, respectively adopting a corresponding noise reduction method to sequentially process the noise pixel points; s3: and outputting the image subjected to noise reduction processing. The corresponding noise reduction method is adopted to process different types of noise pixel points in the input image, so that the detail information of the image can be kept as much as possible, the fuzzy degree of the processed image is greatly reduced, the quality of the image after noise reduction is improved, and the method has the advantages of low algorithm complexity and high applicability.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
It is to be understood that both the foregoing general description and the following detailed description of the present invention are exemplary and explanatory and are intended to provide further explanation of the invention as claimed.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principle of the invention. In the drawings:
FIG. 1 is a flowchart of an image denoising method according to an embodiment of the present invention;
FIG. 2 is a flow chart of an image denoising method according to another embodiment of the present invention;
fig. 3 is a schematic diagram of filter subblocks of size 3 × 3 according to an embodiment of the invention;
fig. 4 is a schematic circuit diagram of an image sensor according to an embodiment of the present invention.
Detailed Description
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the disclosure. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments according to the present application. As used herein, the singular forms "a", "an", and "the" are intended to include the plural forms as well, and it should be understood that when used in this specification the singular forms "a", "an", and/or "the" include "specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
Spatially relative terms, such as "above … …," "above … …," "above … …," "above," and the like, may be used herein for ease of description to describe one device or feature's spatial relationship to another device or feature as illustrated in the figures. It will be understood that the spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. For example, if a device in the figures is turned over, devices described as "above" or "on" other devices or configurations would then be oriented "below" or "under" the other devices or configurations. Thus, the exemplary term "above … …" can include both an orientation of "above … …" and "below … …". The device may be otherwise variously positioned and the spatially relative descriptors used herein interpreted accordingly.
As shown in fig. 1, a first aspect of the present invention provides an image denoising method, including the steps of:
the process first proceeds to step S1: judging whether the input image contains noise pixel points, and if so, further judging the type of each noise pixel point;
then, the process proceeds to step S2: according to the types of the noise pixel points, respectively adopting a corresponding noise reduction method to sequentially process the noise pixel points;
then, the process proceeds to step S3: and outputting the image subjected to noise reduction processing.
In the image noise reduction method, the corresponding noise reduction methods are respectively adopted to process different types of noise pixel points in the input image, so that the detail information on the image can be kept as much as possible, the fuzzy degree of the processed image is greatly reduced, the quality of the image after noise reduction is improved, and the method has the advantages of low algorithm complexity and high applicability.
In some embodiments, the types of the noise pixel points include salt and pepper noise points and/or gaussian noise points. Step S1 includes: and judging whether the input image contains salt and pepper noise points and/or Gaussian noise points. Preferably, in step S1, the number of salt and pepper noise points on the input image is determined, and after performing corresponding filtering (such as median filtering), it is determined whether there is a gaussian noise point on the input image. If the input image has neither salt and pepper noise points nor gaussian noise points, it can be determined that there are no noise pixel points on the current input image, and if any one of the salt and pepper noise points or gaussian noise points exists, it can be determined that there are noise pixel points on the current input image.
In some embodiments, the noise reduction processing on the input image is performed in units of filter sub-blocks, and the size of the filter sub-blocks may be set according to actual needs, and is preferably NXN sub-blocks (N is an odd number), such as 3X3 sub-blocks, 5X5 sub-blocks, and the like.
Judging whether the input image contains salt and pepper noise points comprises the following steps: calculating the average value of all pixel points in a filter subblock with a preset size on an input image; sequentially judging whether the difference between the pixel value of each pixel point in the filtering sub-block and the average value of all the pixel points in the filtering sub-block is greater than a first threshold value or not, and if so, judging that the pixel point in the filtering sub-block is a salt and pepper noise point; otherwise, the pixel is judged to be a salt and pepper noise point.
Further, if it is determined that the filter sub-block contains a salt and pepper noise point, step S2 includes: s21: and according to the difference of the quantity of the salt and pepper noise points in the filter subblocks, performing weighted median filtering processing on each salt and pepper noise point, and taking a pixel value obtained after weighted median filtering processing as an output value of a central pixel point corresponding to the current filter subblock.
Preferably, the "performing weighted median filtering processing on each salt and pepper noise point according to the difference of the number of salt and pepper noise points in the filter sub-block" includes: when the number of the non-salt-and-pepper noise points in the filter subblock is more than two, sorting according to the pixel value of each non-salt-and-pepper noise point, performing weighted operation on each sorted non-salt-and-pepper noise point according to a set weight value, and taking an operation result as an output value of a central pixel point corresponding to the current filter subblock. The weighted value of the non-salt and pepper noise point is in direct proportion to the pixel value of the non-salt and pepper noise point, namely the larger the pixel value of the non-salt and pepper noise point is, the larger the corresponding weighted value is.
Therefore, according to the difference of the quantity of the salt and pepper noise points in the filter subblocks, different filtering strategies are adopted for processing the output value of the central pixel point, so that the image pixels can be kept as much as possible under different conditions, and the image quality is improved.
As shown in fig. 3, the size of the filter subblock is a 3X3 subblock, the 3X3 subblock includes 9 pixels in 3 rows and 3 columns, the pixel values of the pixels in the first row are recorded as C1, C2 and C3 from left to right, the pixel values of the pixels in the second row are recorded as C4, C5 and C6 from left to right, the pixel values of the pixels in the third row are recorded as C7, C8 and C9 from left to right, and the C5 is a central pixel;
step S21 includes:
when all the pixel points in the 3X3 sub-block are judged to be pepper-salt noise points, the calculation formula of the output value q of the central pixel point is as follows: q ═ 5 (C2+ C4+ C5+ C6+ C8);
when all pixel points in the 3X3 sub-block are judged not to be salt and pepper noise points, taking the pixel value of the current central pixel point as the output value q of the central pixel point;
when the number of the non-salt-and-pepper noise points in the 3X3 sub-block is judged to be 1, taking the pixel value of the non-salt-and-pepper noise points as the output value q of the central pixel point;
when the number of the non-salt-and-pepper noise points in the 3X3 sub-block is determined to be 2, the calculation formula of the output value q of the central pixel point is as follows: q ═ Cb × b + Ca × a, Ca < Cb; ca and Cb are the pixel values of two non-salt-pepper noise points respectively; a is the weight value corresponding to the Ca, and b is the weight value corresponding to the Cb. Preferably, a is 0.25 and b is 0.75.
When it is judged thatWhen the number of the non-salt-and-pepper noise points in the 3X3 sub-block is more than 3, the calculation formula of the output value q of the central pixel point is as follows: q ═ C1 × n1+C2*n2+…Cn*nnWherein; C1-Cn is the pixel value of each non-salt-and-pepper noise point, n1-nnAnd C1-Cn are weight values corresponding to the non-salt-pepper noise points. Preferably, in this embodiment, the weight value corresponding to each non-salt-and-pepper noise point is 0.75/s or 0.25/t, s is the number of non-salt-and-pepper noise points with a larger pixel value, and t is the number of non-salt-and-pepper noise points with a smaller pixel value. For example, the number of non-salt and pepper noise points is 3, and the noise points are respectively C1, C2, C3 and C1<C2<C3, then C3 may be regarded as the non-salt-pepper noise point with the larger pixel value in the current filter sub-block, C1 and C2 may be regarded as the non-salt-pepper noise point with the smaller pixel value in the current filter sub-block, and the output value q of the center pixel point is C3 × 0.75+ (C1+ C2) × 0.25/2.
In certain embodiments, the method comprises: after all pixel points in the filtering subblock are processed by the salt and pepper noise point, whether the filtering subblock contains a Gaussian noise point or not is judged, and the method specifically comprises the following steps: and calculating whether the absolute value of the pixel gradient in the filter subblock is smaller than a second threshold value, and if so, judging that the filter subblock contains a Gaussian noise point.
Preferably, the method comprises: when the filter subblocks contain Gaussian noise points, performing weighted mean filtering processing on the output values of the central pixel points obtained by salt and pepper noise processing to obtain the final output values of the central pixel points of the current filter subblocks; the weighted mean filtering process includes: a first weight value is configured for a first pixel array within the filtering sub-block, and a second weight value is configured for a second pixel array within the filtering sub-block.
The final output value of the central pixel point of the current filter subblock is the pixel mean value of all pixel points of the first pixel array, the first weight value, the pixel mean value of all pixel points of the second pixel array, the second weight value and the output value of the central pixel point obtained by salt and pepper noise processing; the first pixel array is a pixel point which is transversely adjacent or longitudinally adjacent to a central pixel point in the current filter subblock, and the second pixel array is a pixel point except the first pixel array and the central pixel point in the current filter subblock.
And when whether the filtering image contains the Gaussian noise point or not is processed, a weighted mean filtering method is adopted. Since weighted mean filtering easily causes blurring of image edge details, edge detection is performed on a central pixel point of a filter subblock before weighted mean filtering is performed, so as to determine whether the central pixel point of a current filter subblock is located at an image edge position. When the central pixel point is located at the edge position of the image, if the central pixel point is still subjected to mean filtering, edge details are easy to blur, so that under the condition, the output value q of the central pixel point of the current filter subblock after salt and pepper noise processing is directly output as a final output value, mean filtering is not performed any more, and the edge details on the image are better reserved.
On the contrary, if the central pixel point is not located at the edge of the image, the output value q after the salt and pepper noise processing is further filtered by adopting a weighted mean filtering mode to obtain the final output value of the central pixel point of the current filtering subblock.
Also taking the 3 × 3 sub-block shown in fig. 3 as an example, the manner of determining whether the current center pixel point is at the edge position is as follows:
first, the absolute value of the gradient of the pixel in the current filter subblock is calculated according to the following formula: j ═ 4 × q-C2-C4-C6-C8|, where J denotes the absolute value of the pixel gradient in the filter subblock, q is the output value of the center pixel point subjected to the salt-pepper filtering processing, and C2, C4, C6, and C8 are the pixel values of the pixel points right above, right left, right, and right below the center pixel point C5 of the current filter subblock in sequence.
Then comparing J with a second threshold value T2, and if J is smaller than T2, indicating that the central pixel point of the filter subblock is not located at the edge position of the image, performing weighted mean filtering processing on the central pixel point; if J is larger than the second T2, it is indicated that the center pixel point of the filter sub-block is located at the edge of the image, and the pixel value obtained after the q is processed by the mean filtering algorithm is used as the final pixel value of the center pixel point of the current filter sub-block and is output.
In this embodiment, the mean filtering algorithm for the center pixel points at the non-edge positions is processed as follows:
d ═ q + (C2+ C4+ C6+ C8) × 0.75+ (C1+ C3+ C7+ C9) × 0.25]/9, D represents the final output value of the center pixel point obtained after the filter subblocks are subjected to weighted mean filtering, and q is the output value q of the center pixel point obtained after the weighted median filtering (i.e., the noise reduction processing is performed on the salt and pepper noise point).
In certain embodiments, the method comprises: and moving the positions of the filtering sub-blocks in the input image according to a preset moving sequence, judging the types of noise pixel points in the new filtering sub-blocks, and performing noise reduction processing by adopting a corresponding noise reduction method until all pixel points on the input image are completely traversed, so as to obtain an image subjected to noise reduction processing and output the image. Preferably, the predetermined moving sequence is scanning line by line from left to right and from top to bottom.
In this embodiment, the first threshold T1 and the second threshold T2 are both related to the neighborhood mean of the filter sub-block. According to the weber theorem, the sensitivity of human eyes to noise is lower in a bright area (an image area with a larger pixel gray scale mean value) than in a dark area (an image area with a smaller pixel gray scale mean value), so that when the central pixel point of the filter subblock is located in the bright area, the threshold value can be properly reduced, and the noise reduction degree is weakened to keep more image edge information; in the dark region, the threshold should be increased as much as possible to suppress noise as much as possible. The degree of brightness of the local area of the image can be characterized by the average value M of the filter sub-blocks, where the value of M is large, the first threshold T1 and the second threshold T2 should be smaller, and where the value of M is small, the first threshold T1 and the second threshold T2 should be larger.
Fig. 2 is a flowchart of an image denoising method according to another embodiment of the present invention. The method comprises the following steps:
the process first proceeds to step S201 to input an image.
And then, step S202 is carried out to judge whether the salt and pepper noise point exists in the current filtering sub-block of the input image, if so, step S203 is carried out to carry out weighted median filtering processing on the central pixel point of the filtering sub-block, otherwise, step S204 is directly carried out to judge whether the Gaussian noise point exists in the current filtering sub-block of the input image.
After step S203, step S204 may also be performed to determine whether there is a gaussian noise point in the current filtering sub-block on the input image, if the determination result of S204 is yes, it indicates that the current center pixel is located at the edge position, and step S207 may be performed to output an image (i.e., the output value q of the center pixel after performing noise reduction processing on the salt and pepper noise point is taken as the final output value); if the determination result of S204 is "no", it indicates that the current center pixel is not located at the edge position, then step S206 is performed to perform weighted mean filtering on the center pixel of the filter sub-block, and the pixel value of the center pixel after weighted filtering is used as the final output value, and then step S207 is performed to output the image.
The second aspect of the present invention also provides a storage medium storing a computer program which, when executed by a processor, performs the method according to the first aspect of the present invention. The storage medium is an electronic component with a data storage function, and includes but is not limited to: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
As shown in fig. 4, a circuit structure of the image sensor according to an embodiment of the present invention is schematically illustrated. The image processor comprises a photosensitive pixel array 301, a signal amplifier 302, an analog-to-digital converter 303, an image data buffer unit 304 and a data output control unit 305, wherein the photosensitive pixel array 301 is connected with the signal amplifier 302, the signal amplifier 302 is connected with the analog-to-digital converter 303, the analog-to-digital converter 303 is connected with the image data buffer unit 304, and the image data buffer unit 304 is connected with the data output control unit 305.
The photosensitive pixel array 301, the signal amplifier 302 and the analog-to-digital converter 303 form an analog noise reduction part, and the image data cache unit 304 is a digital noise reduction processing part.
For reasons of space and limitation, the above description is only a matter of technology that is closely related to the invention of the present application, and some conventional steps for forming an image sensor are not described in detail, but those skilled in the art can combine the present application with the conventional steps of the prior art on the basis of the prior art, and thus the description thereof is omitted here.
From the above description, it can be seen that the above-described embodiments of the present application achieve the following technical effects: the corresponding noise reduction method is adopted to process different types of noise pixel points in the input image, so that the detail information of the image can be kept as much as possible, the fuzzy degree of the processed image is greatly reduced, the quality of the image after noise reduction is improved, and the method has the advantages of low algorithm complexity and high applicability.
It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention. It is therefore intended that the present invention cover the modifications and variations of this invention provided they come within the scope of the appended claims and their equivalents.
Claims (10)
1. An image noise reduction method, comprising the steps of:
s1: judging whether the input image contains noise pixel points, and if so, further judging the type of each noise pixel point;
s2: according to the types of the noise pixel points, respectively adopting a corresponding noise reduction method to sequentially process the noise pixel points;
s3: and outputting the image subjected to noise reduction processing.
2. The image noise reduction method according to claim 1, wherein the types of the noise pixel points include salt-pepper noise points and/or gaussian noise points;
step S1 includes: and judging whether the input image contains salt and pepper noise points and/or Gaussian noise points.
3. The image noise reduction method according to claim 2, wherein determining whether the input image contains salt and pepper noise points comprises:
calculating the average value of all pixel points in a filter subblock with a preset size on an input image;
sequentially judging whether the difference between the pixel value of each pixel point in the filtering sub-block and the average value of all the pixel points in the filtering sub-block is greater than a first threshold value or not, and if so, judging that the pixel point in the filtering sub-block is a salt and pepper noise point; otherwise, the pixel is judged to be a salt and pepper noise point.
4. The image noise reduction method according to claim 3, wherein if it is determined that the filter sub-block contains a salt and pepper noise point, step S2 includes:
s21: and according to the difference of the quantity of the salt and pepper noise points in the filter subblocks, performing weighted median filtering processing on each salt and pepper noise point, and taking a pixel value obtained after weighted median filtering processing as an output value of a central pixel point corresponding to the current filter subblock.
5. The image noise reduction method according to claim 4, wherein performing weighted median filtering processing on the salt and pepper noise points according to the difference of the number of the salt and pepper noise points in the filter sub-block comprises:
when the number of the non-salt-and-pepper noise points in the filter subblock is more than two, sorting according to the pixel value of each non-salt-and-pepper noise point, performing weighted operation on each sorted non-salt-and-pepper noise point according to a set weight value, and taking an operation result as an output value of a central pixel point corresponding to the current filter subblock; the weight value of the non-salt and pepper noise point is in direct proportion to the pixel value of the non-salt and pepper noise point.
6. The image noise reduction method according to claim 4, wherein the size of the filter sub-block is 3X3 sub-block, the 3X3 sub-block includes 9 pixels in 3 rows and 3 columns, the pixel values of the first row of pixels are marked as C1, C2 and C3 from left to right, the pixel values of the second row of pixels are marked as C4, C5 and C6 from left to right, the pixel values of the third row of pixels are marked as C7, C8 and C9 from left to right, and the C5 is a central pixel;
step S21 includes:
when all the pixel points in the 3X3 sub-block are judged to be pepper-salt noise points, the calculation formula of the output value q of the central pixel point is as follows: q ═ 5 (C2+ C4+ C5+ C6+ C8);
when all pixel points in the 3X3 sub-block are judged not to be salt and pepper noise points, taking the pixel value of the current central pixel point as the output value q of the central pixel point;
when the number of the non-salt-and-pepper noise points in the 3X3 sub-block is judged to be 1, taking the pixel value of the non-salt-and-pepper noise points as the output value q of the central pixel point;
when the number of the non-salt-and-pepper noise points in the 3X3 sub-block is determined to be 2, the calculation formula of the output value q of the central pixel point is as follows: q ═ Cb × b + Ca × a, Ca < Cb; ca and Cb are the pixel values of two non-salt-pepper noise points respectively; a is the weight value corresponding to the Ca, and b is the weight value corresponding to the Cb;
when the number of the non-salt-and-pepper noise points in the 3X3 sub-block is judged to be more than 3, the calculation formula of the output value q of the central pixel point is as follows: q ═ C1 × n1+C2*n2+…Cn*nn(ii) a C1-Cn is the pixel value of each non-salt-and-pepper noise point, n1-nnAnd C1-Cn are weight values corresponding to the non-salt-pepper noise points.
7. The image noise reduction method according to claim 4, wherein the method comprises:
after all pixel points in the filtering subblock are processed by the salt and pepper noise point, whether the filtering subblock contains a Gaussian noise point or not is judged, and the method specifically comprises the following steps: and calculating whether the absolute value of the pixel gradient in the filter subblock is smaller than a second threshold value, and if so, judging that the filter subblock contains a Gaussian noise point.
8. The image noise reduction method according to claim 7, comprising:
when the filter subblocks contain Gaussian noise points, performing weighted mean filtering processing on the output values of the central pixel points obtained by salt and pepper noise processing to obtain the final output values of the central pixel points of the current filter subblocks;
the weighted mean filtering process includes: configuring a first weight value for a first pixel array in the filtering sub-block and a second weight value for a second pixel array in the filtering sub-block;
the final output value of the central pixel point of the current filter subblock is the pixel mean value of all pixel points of the first pixel array, the first weight value, the pixel mean value of all pixel points of the second pixel array, the second weight value and the output value of the central pixel point obtained by salt and pepper noise processing;
the first pixel array is a pixel point which is transversely adjacent or longitudinally adjacent to a central pixel point in the current filter subblock, and the second pixel array is a pixel point except the first pixel array and the central pixel point in the current filter subblock.
9. The method of image noise reduction according to claim 8, comprising:
and moving the positions of the filtering sub-blocks in the input image according to a preset moving sequence, judging the types of noise pixel points in the new filtering sub-blocks, and performing noise reduction processing by adopting a corresponding noise reduction method until all pixel points on the input image are completely traversed, so as to obtain an image subjected to noise reduction processing and output the image.
10. A storage medium, characterized in that the storage medium stores a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011618368.6A CN112614078A (en) | 2020-12-30 | 2020-12-30 | Image noise reduction method and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011618368.6A CN112614078A (en) | 2020-12-30 | 2020-12-30 | Image noise reduction method and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112614078A true CN112614078A (en) | 2021-04-06 |
Family
ID=75249527
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011618368.6A Pending CN112614078A (en) | 2020-12-30 | 2020-12-30 | Image noise reduction method and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112614078A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114037635A (en) * | 2021-11-26 | 2022-02-11 | 成都微光集电科技有限公司 | An image hybrid denoising method, device, electronic device and medium |
| CN115358951A (en) * | 2022-10-19 | 2022-11-18 | 广东电网有限责任公司佛山供电局 | Intelligent ring main unit monitoring system based on image recognition |
| CN120580165B (en) * | 2025-08-05 | 2025-10-14 | 中国石油大学(华东) | A method for image classification denoising based on image edge noise characteristics |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020911A (en) * | 2012-12-12 | 2013-04-03 | 中国航空工业集团公司洛阳电光设备研究所 | Image processing method for filtering mixed noise |
| CN103414845A (en) * | 2013-07-24 | 2013-11-27 | 中国航天科工集团第三研究院第八三五七研究所 | Self-adaptive video image noise reducing method and noise reducing system |
| US20150104100A1 (en) * | 2013-01-07 | 2015-04-16 | Huawei Device Co., Ltd. | Image Processing Method and Apparatus, and Shooting Terminal |
| CN105260991A (en) * | 2015-09-29 | 2016-01-20 | 中国电子科技集团公司第四十四研究所 | Adaptive hybrid noise reduction method for CMOS image sensor noise reduction processing |
| CN111681174A (en) * | 2020-04-29 | 2020-09-18 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Joint Filtering Method for Suppressing Salt and Pepper/Gaussian Mixed Noise Target Image |
-
2020
- 2020-12-30 CN CN202011618368.6A patent/CN112614078A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103020911A (en) * | 2012-12-12 | 2013-04-03 | 中国航空工业集团公司洛阳电光设备研究所 | Image processing method for filtering mixed noise |
| US20150104100A1 (en) * | 2013-01-07 | 2015-04-16 | Huawei Device Co., Ltd. | Image Processing Method and Apparatus, and Shooting Terminal |
| CN103414845A (en) * | 2013-07-24 | 2013-11-27 | 中国航天科工集团第三研究院第八三五七研究所 | Self-adaptive video image noise reducing method and noise reducing system |
| CN105260991A (en) * | 2015-09-29 | 2016-01-20 | 中国电子科技集团公司第四十四研究所 | Adaptive hybrid noise reduction method for CMOS image sensor noise reduction processing |
| CN111681174A (en) * | 2020-04-29 | 2020-09-18 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Joint Filtering Method for Suppressing Salt and Pepper/Gaussian Mixed Noise Target Image |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114037635A (en) * | 2021-11-26 | 2022-02-11 | 成都微光集电科技有限公司 | An image hybrid denoising method, device, electronic device and medium |
| CN114037635B (en) * | 2021-11-26 | 2025-06-27 | 成都微光集电科技有限公司 | Image mixed denoising method, device, electronic device and medium |
| CN115358951A (en) * | 2022-10-19 | 2022-11-18 | 广东电网有限责任公司佛山供电局 | Intelligent ring main unit monitoring system based on image recognition |
| CN120580165B (en) * | 2025-08-05 | 2025-10-14 | 中国石油大学(华东) | A method for image classification denoising based on image edge noise characteristics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2037066C (en) | Image processing technique | |
| CN110796615B (en) | Image denoising method, device and storage medium | |
| CN114119436B (en) | Infrared image and visible light image fusion method, device, electronic equipment and medium | |
| US20100189374A1 (en) | Image processor and program | |
| US20100002953A1 (en) | Detection and reduction of ringing artifacts based on block-grid position and object edge location | |
| KR20110002858A (en) | Filtering method and apparatus for anti-aliasing | |
| CN110246089A (en) | Bayer area image noise reduction system and its method based on non-local mean filter | |
| Toh et al. | Cluster-based adaptive fuzzy switching median filter for universal impulse noise reduction | |
| CN112614078A (en) | Image noise reduction method and storage medium | |
| CN104899842B (en) | The adaptive extreme value median filter method of sequence for remote line-structured light image | |
| US20090074318A1 (en) | Noise-reduction method and apparatus | |
| Ma et al. | An automatic detection method of Mura defects for liquid crystal display | |
| CN111242884B (en) | Image dead pixel detection and correction method and device, storage medium and camera equipment | |
| CN115311155B (en) | Improved KPN network image deraining method, system and storage medium | |
| Han et al. | Smear removal algorithm using the optical black region for CCD imaging sensors | |
| WO2021189460A1 (en) | Image processing method and apparatus, and movable platform | |
| Vijaykumar et al. | Decision based adaptive median filter to remove blotches, scratches, streaks, stripes and impulse noise in images | |
| CN116152186B (en) | Submerged plant coverage area intelligent monitoring method | |
| CN120318118B (en) | Video deblurring method, medium and device based on 3D spatiotemporal grid perception | |
| Kaur et al. | An improved weighted median filter for the image processing application | |
| CN108133475B (en) | Detection method of local focus blurred image | |
| CN119338686B (en) | A method for suppressing non-uniform background of wide-area star images using a CCD sensor | |
| CN118485580B (en) | Image enhancement method and device based on bilateral filtering, equipment and storage medium | |
| Hota et al. | Removal of random valued impulse noise with modified weighted mean filter by using denoising algorithm: Emerging opportunities of image processing application | |
| CN120339100A (en) | Image enhancement method based on ISP image signal processing visual sensor |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB02 | Change of applicant information |
Address after: Office Buildings 1 and 5, Phase I, Optoelectronic Information Industry Park, No. 7691 Ziyou Road, Changchun Economic and Technological Development Zone, Jilin Province, 130000 Applicant after: Changchun Changguang Chenxin Microelectronics Co.,Ltd. Address before: No. 588, Yingkou Road, Jingkai District, Changchun City, Jilin Province, 130033 Applicant before: Changchun Changguangchenxin Optoelectronics Technology Co.,Ltd. |
|
| CB02 | Change of applicant information |