JP2007049241A - Crosstalk correction device and method for scanner - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a crosstalk correction device for scanner which can also cope with variation in scanning conditions incident to difference in density or reflectivity of a document or deterioration of a light source. <P>SOLUTION: In the preprocssing stage of scanning, a color CCD 7 reads in a white reference surface 4 and a black line portion 5 to predetermine the reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the amount of crosstalk imparted from R (red), G (green) and B (blue) of one line to G (green) and B (blue), R (red) and B (blue), and R (red) and G (green) of each other line. When scanning is performed, image data R2, G2 and B2 of each line is corrected to eliminate crosstalk in real time based on the image data R2, G2 and B2 of each line read in by scanning a document 9 and the reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the amount of crosstalk corresponding to each other line. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a crosstalk correction device for a scanner and a crosstalk correction for correcting crosstalk between lines which causes a problem when a document is read by a color CCD having different light receiving characteristics provided with a certain line interval. Regarding the method.

  Conventionally, when a color image is read by a line CCD, a method using a reading line for three colors of R (red) / G (green) / B (blue) on one chip has been used. The CCD reading lines have a distance of several lines between the lines, and when scanning scanning, the three reading lines do not read the same position on the document surface at the same time. Therefore, the inter-line correction process is performed by the inter-line correction circuit in consideration of the delay between the lines.

  Also, as a characteristic of the CCD, there is a problem that a crosstalk of a certain level (usually about 1% to 2%) occurs with respect to other lines according to the level of the pixel read at the same time. The effect of this crosstalk causes a phenomenon such as an afterimage in the image after line-to-line correction, and the afterimage becomes more conspicuous as the line spacing is larger, and the color balance is greatly lost, which adversely affects the image quality. It is easy to come out.

  As a solution to the above problem, the line spacing is shortened as much as possible, and even if crosstalk occurs, the influence on the image quality is made as inconspicuous as possible, and the influence increases due to variations in the maximum output voltage for each line. As disclosed in Document 1 and the like, there has been considered a method of suppressing crosstalk by suppressing the variation in sensitivity for each line by reducing the number of lines and adjusting the cell size.

However, as reported in Patent Document 2, in order to shorten the line interval, the light receiving part of the CCD for three lines is brought close and the charge transfer part (shift register) is accumulated when the distance is increased. When the charge is sent to the shift register, it is affected by charges of different colors, and crosstalk occurs again here. In the method according to Patent Document 2, a document to be used for correction must be read twice, which may cause a problem in terms of scanner productivity.
In addition to reducing the number of lines, a special mirror is required in addition to the normal mirror, and when the line spacing is shortened as much as possible, the distance from the photoconductor to the shift register may be unbalanced. The adverse effect on the image quality is considered due to
Further, when developing a dedicated color CCD such as adjusting the cell size without using a conventional color CCD, the scale of development is large and the cost increases.

On the other hand, as a countermeasure against the occurrence of crosstalk when a general 3-line CCD is used, a value obtained by multiplying the image signal of the mixed color line by the correction coefficient is read and subtracted from the image signal of the image line. A technical idea of acquiring an image signal is proposed in Patent Document 3.
However, the correction coefficient is only a fixed value obtained in advance by simulation, and it is difficult to dynamically cope with changes in scanning conditions due to differences in document density and reflectance, or deterioration of the light source. The problem remains.

JP-A-7-322006 JP-A-8-65512 (paragraph number 0006) JP-A-11-331624 (paragraph numbers 0016, 0031)

  Accordingly, an object of the present invention is to improve the inconveniences of the above-described prior art, cope with changes in scanning conditions due to differences in document density and reflectance, deterioration of light sources, etc. Another object of the present invention is to provide a scanner crosstalk correction device and a crosstalk correction method capable of accurately suppressing crosstalk between the two.

The scanner crosstalk correction apparatus according to the present invention is a scanner crosstalk for correcting crosstalk generated between lines when a document is read by a color CCD having different light receiving characteristics provided with a fixed line interval. In order to achieve the above-mentioned problem,
Crosstalk amount calculating means for reading a white reference and a black reference by the color CCD at a pre-processing stage for scanning a document and obtaining a reference value of a crosstalk amount that one line gives to each other line for each line; ,
A cross that applies correction for eliminating crosstalk in real time to the image data of each line based on the image data of each line scanned and read from the original and the reference value of the crosstalk amount corresponding to each other line. And a talk correction output means.

  At the pre-processing stage of scanning the original, the white reference and the black reference are read by the color CCD, and a reference value of the crosstalk amount that one line gives to each other line is obtained in advance. The image data of each line is corrected in real time to eliminate crosstalk based on the image data of each line scanned by scanning the document and the reference value of the crosstalk amount corresponding to each other line. Therefore, it is possible to dynamically cope with changes in scanning conditions due to differences in document density and reflectance, or deterioration of the light source, and to accurately suppress crosstalk between lines using a general-purpose three-line CCD. I can do it now.

As a more specific configuration, it comprises a white reference composed of a white reference surface disposed on a contact glass for placing an original, and a black reference composed of a black line portion provided in the white reference surface,
The crosstalk amount calculating means locates the white reference image data of each line read by positioning the color CCD of each line on the white reference plane, and positions one of the lines on the black line portion. Based on the deviation between the read white image data of the other lines and the white reference image data of the other lines, the reference value of the crosstalk amount given to the other lines by the one line is determined for each line. Seeking
The crosstalk correction output means determines that each deviation between the image data of each line scanned and read from the document and the image data and the image data of each other line other than the image data is the maximum density expression. A configuration is proposed in which correction for eliminating crosstalk is performed on the image data of each line on the basis of the ratio occupied by the line and the reference value of the crosstalk amount corresponding to each other line.

For example, when color CCDs having different light receiving characteristics have three colors of R (red) / G (green) / B (blue), the crosstalk amount calculation means firstly R (red) / G (green) / B The white CCD image data R1, G1, B1 of each line is obtained by positioning the color CCD of each line of (blue) on the white reference plane.
Next, one of the lines, for example, the R (red) line is positioned in the black line portion, and the other lines, namely, the G (green) line and the B (blue) line are The white image data Gr, Br is obtained, and the deviations Gr-G1, Br-B1 between the white reference image data G1, B1 of the other lines, that is, the G (green) line and the B (blue) line are obtained. , These values are used as the reference value ΔGr, the crosstalk amount reference value ΔGr, that the R (red) line gives to each of the other lines, that is, the G (green) line and the B (blue) line. Let it be ΔBr. That is, the reference value ΔGr of the crosstalk amount given to the G (green) line by the R (red) line is ΔGr = Gr−G1, and the cross value given to the B (blue) line by the R (red) line The reference value ΔBr of the talk amount is ΔBr = Br−B1.
Similarly to the above, the other one of the lines, for example, the G (green) line is positioned in the black line portion, and the other lines, in other words, the R (red) line. And B (blue) line white image data Rg, Bg is obtained, and the deviation Rg between the other lines, that is, the R (red) line and the B (blue) line white reference image data R1, B1. -R1, Bg-B1 and calculate these values as the other one line, that is, the G (green) line is the other lines, that is, the R (red) line and the B (blue) line. Are the reference values ΔRg and ΔBg of the amount of crosstalk given to. That is, the reference value ΔRg of the crosstalk amount that the G (green) line gives to the red (R) line is ΔRg = Rg−R1, and the cross that the G (green) line gives to the B (blue) line The reference value ΔBg of the talk amount is ΔBg = Bg−B1.
Further, in the same manner as described above, the last one of the lines, for example, the B (blue) line is positioned in the black line portion, and the other lines, in other words, the R (red) line, The white image data Rb, Gb of the G (green) line is obtained, and the deviation Rb− between the other lines, that is, the R (red) line and the white reference image data R1, G1 of the G (green) line is obtained. R1, Gb-G1 is obtained, and these values are changed to the last one line, that is, the B (blue) line is changed to each of the other lines, that is, the R (red) line and the G (green) line. The reference values ΔRb and ΔGb of the crosstalk amount to be given are used. That is, the reference value ΔRb of the crosstalk amount given to the red (R) line by the B (blue) line is ΔRb = Rb−R1, and the cross value given to the G (green) line by the B (blue) line The reference value ΔGb for the talk amount is ΔGb = Gb−G1.
When the document is scanned, the crosstalk correction output means scans the document and reads each line of image data, for example, R (red) line image data R2, and the image data R2. Except for each other line, in short, each deviation R2-G2, R2-B2 between the image data G2, B2 of the G (green) line and B (blue) line is relative to the maximum value P of the density expression. The image of the R (red) line based on the ratio and the reference values ΔGr, ΔBr of the crosstalk amount corresponding to the other lines, in other words, the G (green) line and the B (blue) line Apply corrections to eliminate crosstalk on the data. That is, if the image data after correction of the R (red) line is R3, R3 = R2−ΔGr (R2−G2) / P−ΔBr (R2−B2) / P. In other words, the reference value ΔGr of the crosstalk amount is the image data G1 of the G (green) line read by positioning each R (red) / G (green) / B (blue) line on the white reference plane. Deviation Gr-G1 from the image data Gr of the G (green) line when the R (red) line is positioned on the black line portion and the G (green) line is positioned on the white reference plane. Since it is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of expression, the density deviation R2-G2 of the actually scanned image data and the maximum value P of the density expression Therefore, the weighting of (R2−G2) / P is applied to the reference value ΔGr of the crosstalk amount, so that the influence according to the difference in the density and reflectance of the document or the deterioration of the light source is obtained. The reference value ΔBr of the amount of crosstalk is the image data B1 of the B (blue) line read with the R (red) / G (green) / B (blue) lines all positioned on the white reference plane. Deviation Br-B1 from the image data Br of the B (blue) line when the R (red) line is positioned on the black line portion and the B (blue) line is positioned on the white reference plane, in other words, the density Since it is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of expression, the density deviation R2-B2 of the actually scanned image data and the maximum value P of the density expression Therefore, by applying a weight of (R2-B2) / P to the reference value ΔBr of the crosstalk amount, an influence according to a difference in document density or reflectance, deterioration of a light source, or the like is required. Therefore, finally, subtraction processing R2-ΔGr (R2-G2) / which removes the influence of ΔGr (R2-G2) / P and ΔBr (R2-B2) / P from the image data R2 of the R (red) line. By performing P-ΔBr (R2-B2) / P, it is possible to obtain appropriate image data R3 as R (red) line image data regardless of differences in document density, reflectance, light source deterioration, etc. Can do.
In the same manner as described above, the image data G2 of the G (green) line, the image data G2 and other lines other than the image data G2, that is, the R (red) line and the B (blue) line. The ratio of each deviation G2-R2, G2-B2 between the image data R2, B2 to the maximum value P of the density expression and the other lines, in short, the R (red) line and the B (blue) ) Is corrected for eliminating crosstalk on the image data of the green (G) line based on the reference values ΔRg and ΔBg of the crosstalk amount corresponding to the line. That is, if the image data after correction of the G (green) line is G3, G3 = G2−ΔRg (G2−R2) / P−ΔBg (G2−B2) / P. In other words, the reference value ΔRg of the crosstalk amount is the R (red) line image data R1 read with the R (red) / G (green) / B (blue) lines all positioned on the white reference plane. Deviation Rg-R1 from the image data Rg of the R (red) line when the G (green) line is positioned at the black line portion and the red (R) line is positioned at the white reference plane. Since it is a value obtained corresponding to the gradation difference P between white and black which is the maximum value of expression, the density deviation G2-R2 of the actually scanned image data and the maximum value P of the density expression Therefore, by applying a weighting of (G2-R2) / P to the reference value ΔRg of the crosstalk amount, an influence according to a difference in the density or reflectance of the document or deterioration of the light source is required. The reference value ΔBg for the amount of crosstalk is the image data B1 of the B (blue) line read by positioning each R (red) / G (green) / B (blue) line on the white reference plane. Deviation Bg-B1 from the image data Bg of the B (blue) line when the green (G) line is positioned on the black line and the B (blue) line is positioned on the white reference plane. Since it is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of expression, the density deviation G2-B2 of the actually scanned image data and the maximum value P of the density expression Thus, the weighting of (G2−B2) / P is applied to the reference value ΔBg of the crosstalk amount, so that the influence according to the difference in the density or reflectance of the document or the deterioration of the light source is obtained. Therefore, finally, a subtraction process G2-ΔRg (G2-R2) / that removes the influence of ΔRg (G2-R2) / P and ΔBg (G2-B2) / P from the image data G2 of the G (green) line. By performing P-ΔBg (G2-B2) / P, it is possible to obtain appropriate image data G3 as image data of the G (green) line regardless of differences in the density and reflectance of the original or deterioration of the light source. Can do.
Further, in the same manner as described above, the image data B2 of the B (blue) line and the image data B2 and other lines other than the image data B2, that is, the R (red) line and the G (green) line The ratio of each deviation B2-R2, B2-G2 to the maximum value P of the density expression between the image data R2, G2 and the other lines, in short, the R (red) line and the G (green) ) Is applied to the image data of the blue (B) line based on the reference values ΔRb and ΔGb of the crosstalk amount corresponding to the line). That is, if the image data after correction of the blue (B) line is B3, B3 = B2−ΔRb (B2−R2) / P−ΔGb (B2−G2) / P. In other words, the reference value ΔRb of the crosstalk amount is the image data R1 of the R (red) line read by positioning each R (red) / G (green) / B (blue) line on the white reference plane. Deviation Rb-R1 from the image data Rb of the R (red) line when the B (blue) line is positioned on the black line portion and the red (R) line is positioned on the white reference plane. Since it is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of the expression, the density deviation B2-R2 of the actually scanned image data and the maximum value P of the density expression Therefore, the weighting of (B2−R2) / P is applied to the reference value ΔRb of the crosstalk amount, so that the influence according to the difference in the density and reflectance of the document or the deterioration of the light source is obtained. The reference value ΔGb of the crosstalk amount is the image data G1 of the G (blue) line read by positioning each R (red) / G (green) / B (blue) line on the white reference plane. Deviation Gb-G1 from the image data Gb of the G (green) line when the blue (B) line is positioned at the black line portion and the G (green) line is positioned at the white reference plane. Since it is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of expression, the density deviation G2-B2 of the actually scanned image data and the maximum value P of the density expression Therefore, the weighting of (B2−G2) / P is applied to the reference value ΔGb of the crosstalk amount, so that the influence according to the difference in the density or reflectance of the document or the deterioration of the light source is obtained. Therefore, finally, a subtraction process B2-ΔRb (B2-R2) / which removes the influence of ΔRb (B2-R2) / P and ΔGb (B2-G2) / P from the image data B2 of the B (blue) line. By performing P-ΔGb (B2-G2) / P, it is possible to obtain appropriate image data B3 as the image data of the B (blue) line regardless of differences in the density and reflectance of the original or deterioration of the light source. Can do.

  Here, it is desirable that the black line portion is formed narrower than the width between the lines of the color CCD.

  By forming the black line portion to be narrower than the width between the lines of the color CCD, interference between the lines can be prevented and the reference value of the crosstalk amount can be obtained appropriately.

  The crosstalk amount calculating means temporarily stores the white image data in an inter-line correction memory of an inter-line correction circuit used for delay adjustment of image data for each line, and uses an average value of each white image data. It is desirable to configure so as to obtain a reference value for the amount of crosstalk.

  Even when the number of CCDs constituting each line is large, white image data can be temporarily stored using the inter-line correction memory, and the average value of the white image data of each line can be obtained. Storage capacity can be saved.

Alternatively, in place of the above configuration, a black reference composed of a black reference surface disposed on a contact glass for placing an original, and a white reference composed of a white line portion provided in the black reference surface,
The crosstalk amount calculation means reads the black reference image data of each line read by positioning the color CCD of each line on the black reference plane and one of the lines positioned at the white line portion. On the basis of the deviation between the black image data of each other line and each black reference image data of the other line, the reference value of the crosstalk amount given to each other line by the one line is determined for each line. Seeking
The crosstalk correction output means determines that each deviation between the image data of each line scanned and read from the document and the image data and the image data of each other line other than the image data is the maximum density expression. In other words, the image data of each line may be corrected to eliminate crosstalk on the basis of the ratio occupied by the line and the reference value of the crosstalk amount corresponding to each other line.

Substantial operational effects are the same as those using the white reference composed of the white reference plane and the black reference composed of the black line portion.
However, in this case, since the density reference is reversed from the above, the image after correction of each line of R (red) / G (green) / B (blue) in the actual scanning process is the same as described above. Data is R3 = R2-ΔGr (R2-G2) / P-ΔBr (R2-B2) / P, G3 = G2-ΔRg (G2-R2) / P-ΔBg (G2-B2) / P, B3 = B2- ΔGr = G1-Gr, ΔBr = B1-Br, ΔRg = R1-Rg, ΔBg = B1-Bg, ΔRb = R1 if calculated as ΔRb (B2-R2) / P-ΔGb (B2-G2) / P -Rb, ΔGb = G1-Gb, so that the sign is inverted at the preprocessing stage to obtain the reference value of the crosstalk amount.

  When such a configuration is applied, it is desirable that the white line portion be formed narrower than the width between the lines of the color CCD.

  By forming the white line portion narrower than the width between the lines of the color CCD, interference between lines can be prevented and the reference value of the crosstalk amount can be obtained appropriately.

  Further, the crosstalk amount calculating means is preferably configured to offset black correction so that the black reference image data does not become a value near the limit of density expression.

  Black reference image data R1, G1, B1 read with the color CCD of each line positioned on the black reference plane, and black image data of each other line read with one of the lines positioned at the white line Gr, Br, Rg, Bg, Rb, Gb, that is, the black image data Gr, Br, Rg, Bg, Rb, Gb of each line affected by crosstalk can be surely kept within the range of density expression. Even when the density reference is reversed, the crosstalk can be suppressed by accurately obtaining the reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the crosstalk amount.

  Similarly to the above, the crosstalk amount calculating means temporarily stores the black image data in the inter-line correction memory of the inter-line correction circuit used for delay adjustment of the image data for each line, and averages each black image data It is desirable that the reference value of the crosstalk amount be obtained using

  Even when the number of CCDs constituting each line is large, the black image data can be temporarily stored using the inter-line correction memory, and the average value of the black image data of each line can be obtained. Storage capacity can be saved.

The crosstalk correction method of the present invention is a crosstalk correction method for correcting crosstalk generated between lines when a document is read with a color CCD having different light receiving characteristics provided side by side with a fixed line interval. In order to achieve the same problem as above,
In the pre-processing stage of scanning the document, the white CCD image data of each line read with the color CCD of each line positioned on the white reference and one of the lines positioned on the black reference are read. On the basis of the deviation between the white image data of each other line and the respective white reference image data of the other line, the reference value of the crosstalk amount given to each of the other lines by the one line is determined for each line. Asking
When scanning an original, the ratio of each deviation between the image data of each line read in the scan and the image data and the image data of each of the other lines excluding this to the maximum value of the density expression, The image data of each line is corrected in real time to eliminate crosstalk based on the reference value of the crosstalk amount corresponding to each of the other lines.

  The substantial action is as described above, and dynamically copes with changes in scanning conditions due to differences in document density and reflectance, or deterioration of the light source, and a cross-line between lines using a general-purpose three-line CCD. Talk can be suppressed accurately.

Alternatively, instead of the above configuration, in the pre-processing stage of scanning the document, the black CCD image data of each line read by positioning the color CCD of each line with the black reference and one of the lines are The amount of crosstalk that the one line gives to each other line based on the deviation between the black image data of each other line that is read while being positioned at the white reference and each black reference image data of the other line Obtain a reference value for each line,
When scanning an original, the ratio of each deviation between the image data of each line read in the scan and the image data and the image data of each of the other lines excluding this to the maximum value of the density expression, A configuration may be adopted in which correction for eliminating crosstalk is performed in real time on the image data of each line based on the reference value of the crosstalk amount corresponding to each of the other lines.

  Similar to the above, it is possible to dynamically cope with changes in scanning conditions due to differences in document density and reflectance, light source deterioration, etc., and to accurately suppress crosstalk between lines using a general-purpose three-line CCD. it can.

  The crosstalk correction device and the crosstalk correction method for a scanner according to the present invention read a white reference and a black reference by a color CCD in a preprocessing stage for scanning a document, and a crosstalk amount that one line gives to each other line. In addition, when scanning a document, each line is determined based on the image data of each line scanned and read and the reference value of the crosstalk amount corresponding to each other line. Since correction for eliminating crosstalk is applied to the image data in real time, it is possible to dynamically cope with changes in scanning conditions due to differences in document density and reflectance, light source deterioration, etc. Crosstalk between lines can be accurately suppressed using a line CCD.

  Next, the best mode for carrying out the present invention will be described with reference to the drawings. FIG. 1 is a conceptual diagram showing an outline of the structure of a mechanism part of a scanner crosstalk correction apparatus according to an embodiment to which the crosstalk correction method of the present invention is applied. FIG. 2 is a control of the scanner crosstalk correction apparatus. It is the functional block diagram which showed the outline of the structure of a part.

  As shown in FIG. 1, the scanner crosstalk correction device 1 according to this embodiment is substantially integrated into a scanner 2, and the main part of the mechanism places a document 9 on the scanner 2. White reference planes (white reference) 4 and 6 installed on the contact glass 3 and a black line portion (black reference) 5 provided on the same plane so as to be sandwiched between the white reference planes 4 and 6 Consists of.

  Reference numeral 7 denotes a three-line color CCD provided in the scanner 2, and reference numeral 8 denotes a lamp as a light source. The scanner 2 includes a mirror, a lens, and the like in addition to these, but since all are known elements, description thereof is omitted here. In FIG. 1, the left-right direction is the sub-scanning direction of the scanner 2, and the direction orthogonal to the paper surface is the main-scanning direction, and the color CCD 7 moves in the sub-scanning direction and reads an image from the document 9 on the contact glass 3. The color CCD 7 includes a number of CCDs for each line along the main scanning direction.

  The lamp 8 is a white lamp such as a xenon tube that is tuned with respect to the sensitivity characteristics of the color CCD 7.

  The white reference planes 4 and 6 and the black line portion 5 have a sufficient width with respect to the main scanning direction. Of these, the white reference plane 4 and the white reference plane 6 are obtained when an image is read by the color CCD 7 of the scanner 2. In order to obtain the same value, the same material and the same whiteness are used. Since the white reference surface 4 and the white reference surface 6 have substantially the same characteristics, it can be said that the black line portion 5 is provided within the white reference surface that is entirely constituted by the white reference surfaces 4 and 6. .

  Further, the width of the black line portion 5 in the sub-scanning direction is formed narrower than the width between the lines of the color CCD 7. Therefore, for example, if the color CCD 7 is constituted by a CCD having an interval of 8 lines and the resolution in the sub-scanning direction of the scanner 2 is 600 dpi, the width of the black line portion 5 is 25.4 mm / 600 × (8-1). ) ≈0.3mm, and it is necessary to make the width less than 0.3mm.

  Although the color CCD 7 is not particularly limited with respect to the number of lines, the following description will be made taking an example of a three-line color CCD of R (red) / G (green) / B (blue). Further, there is no particular restriction on the line interval.

  As shown in FIG. 2, the control unit 10 that is a part of the scanner crosstalk correction device 1 includes an image input unit 11, an analog front end unit (AFE) 12, an image processing LSI 13, and a crosstalk correction memory. 14 and an inter-line correction memory 15.

  The image input unit 11 outputs image data read by the color CCD 7 of the scanner 2 shown in FIG. 1 as an analog signal. When there are a plurality of lines (for example, three lines of R (red) / G (green) / B (blue)) in the package in the color CCD 7, signals output at the same time are spaced at intervals of each line. Since the data is read at a certain position, a gap between lines is always generated. The analog signal output from the image input unit 11 is a signal before correcting the gap between lines.

  The analog front end unit (AFE) 12 includes a gain control circuit 16, an A / D conversion circuit 17, and a black correction circuit 18. The gain control circuit 16 has a function of enlarging / reducing the input analog signal at a preset magnification. The black correction circuit 18 has a function of applying an offset to the analog signal so that a desired value is obtained when black is read. The A / D conversion circuit 17 has a function of converting an input analog signal into a digital signal having a predetermined number of bits and outputting the digital signal.

  In the following description, it is assumed that the output of the A / D conversion circuit 17 is 8-bit 256 gradation, pure white corresponds to the numerical value 255, and jet black corresponds to the numerical value 0. That is, the maximum value of the density expression is 256.

The image processing LSI 13 includes a crosstalk correction circuit 19 and an interline correction circuit 20.
The crosstalk correction circuit 19 once stores the image data read by the color CCD 7 in the inter-line correction memory 15 and reads the image data from the inter-line correction memory 15 at the stage of preprocessing for scanning the document 9. A function as a crosstalk amount calculating means for calculating a reference value of the crosstalk amount for each line, a function of storing the obtained reference value of the crosstalk amount in the crosstalk correction memory 14, and a scan of the document 9. As a crosstalk correction output means for reading a reference value of the amount of crosstalk from the memory 14 for crosstalk correction sometimes and applying correction for eliminating crosstalk to the input image data on the basis of the reference value. It has the function of.
Further, the inter-line correction circuit 20 stores image data corresponding to the line interval × (number of lines−1) matching the specifications of the color CCD 7 in the inter-line correction memory 15 and delays the image data for each line. It has a function of performing adjustment and generating image data so that a signal from the image input unit 11 including a position gap becomes data read at the same position at the same time.

  The scanner crosstalk correction device 1 according to this embodiment is configured such that the reference value of the crosstalk amount for each line of R (red) / G (green) / B (blue) at the stage of preprocessing for starting scanning of the document 9. Is calculated and stored in the crosstalk correction memory 14, and when the document 9 is scanned, the crosstalk generated between the lines is corrected by using the reference value of the crosstalk amount. The data of each line from which the influence of the talk is removed is corrected so as to become data obtained by reading the same position at the same time by the inter-line correction processing similar to the conventional one.

  First, the processing of the crosstalk amount calculation means (crosstalk correction circuit 19) related to the calculation of the reference value of the crosstalk amount will be described with reference to the flowchart of FIG.

  First, the lamp 8 is turned on (step a1), and a gain of 1 is set for the gain control circuit 16 of the analog front end unit (AFE) 12 (step a2).

  Next, the line of sight of the color CCD 7 is moved and stopped just below the white reference plane 4 (step a3, step a4). The stop position is in the vicinity of the center of the white reference plane 4 so that the three lines of the color CCD 7 are stable and the white of the white reference plane 4 can be read.

  A weight is applied until the light quantity of the lamp 8 is stabilized, and the color CCD 7 reads white on the white reference surface 4 (step a5).

  For each of R (red) / G (green) / B (blue), the reading range in the main scanning direction where the white reference appears brightest is set as the reading start position PIXS and end position PIXE, and the range The average value of the image data of each line is obtained for each of R (red) / G (green) / B (blue), and each of the obtained R (red) / G (green) / B (blue) is obtained. The average value is set as the white reference image data R1, G1, B1, and is stored in the crosstalk correction memory 14 together with the above-described PIXS, PIXE (step a6).

  Next, a movement amount corresponding to the number of sub-scanning lines necessary for reading the black line portion 5 is set (step a7), and each line of R (red) / G (green) / B (blue) is set. The movement of the line of sight of the color CCD 7 is started so as to read the black line portion 5.

  The image data read here is stored in the inter-line correction memory 15 by the crosstalk correction circuit 19 of the image processing LSI 13. This reading operation and the storing operation in the inter-line correction memory 15 are repeated until the set number of sub-scanning lines is read (step a8 to step a10).

  Thereafter, the eye movement of the color CCD 7 is stopped (step a11), and the reference value of the crosstalk amount is calculated (step a12).

  A method for calculating the reference value of the crosstalk amount will be described in detail below.

  First, image data when the black line portion 5 is read is read from the inter-line correction memory 15 of the image processing LSI 13.

  Then, the average values R0, G0, B0 in the above-mentioned reading range PIXS to PIXE are calculated for each of the read R (red) / G (green) / B (blue) data. Save in the correction memory 15. This calculation of the average value is repeated for the number of sub-scanning lines obtained by reading the black line portion 5.

  When the contents of the average values R0, G0, B0 stored in the interline correction memory 15 are viewed in time series, a waveform as shown in FIG. 4 is obtained.

  Here, an 8-line color CCD 7 is taken as an example, and the width of the black line portion 5 is described as 7 lines.

  First, an average value for 7 lines is obtained for the pixel levels of the remaining G (green) line and B (blue) line when the R (red) line reads the black line portion 5 and the value changes.

  The obtained average values are set as white image data Gr and Br, respectively.

Here, the image data when the white reference plane 4 is read in advance (see step a6) is set as white reference image data R1, G1, B1, and the amount of crosstalk from the R (red) line to the G (green) line is determined. When the reference value is ΔGr and the reference value of the crosstalk amount from the R (red) line with respect to the B (blue) line is ΔBr, the reference values ΔGr and ΔBr of the crosstalk amount are expressed by the following equations (1) and (2 ).
ΔGr = Gr-G1 (1)
ΔBr = Br-B1 (2)

Similarly, the average values of seven lines of the R (red) line and the B (blue) line when the G (green) line is changed by reading the black line portion 5 are set as white image data Rg and Bg, respectively. At this time, if the reference value of the crosstalk amount from the G (green) line for the R (red) line is ΔRg and the reference value of the crosstalk amount from the G (green) line for the B (blue) line is ΔBg, The reference values ΔRg and ΔBg of the talk amount are obtained by the following equations (3) and (4).
ΔRg = Rg-R1 (3)
ΔBg = Bg-B1 (4)

Similarly, when the B (blue) line is changed by reading the black line portion 5, the average values for the seven lines of the R (red) line and the G (green) line are respectively obtained as the white image data Rb, Gb. And At this time, if the reference value of the crosstalk amount from the B (blue) line to the R (red) line is ΔRb and the crosstalk amount from the B (blue) line to the G (green) line is ΔGb, the crosstalk amount is The reference values ΔRb and ΔGb are obtained by the following equations (5) and (6).
ΔRb = Rb-R1 (5)
ΔGb = Gb-G1 (6)

  After calculating the reference values of the crosstalk amounts in the above formulas (1) to (6), the crosstalk amount reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb for the obtained lines are crosstalk corrected. (Step a13), and the calculation process related to the reference value of the crosstalk amount is terminated.

  Then, the color CCD 7 is moved to the white reference plane 6 and a desired gain is applied to the analog signal from the color CCD 7 by the gain control circuit 16 of the analog front end unit (AFE) 12. A shading correction calculation is performed to compensate for non-uniformity in the amount of light.

  Next, processing of a crosstalk correction output means (crosstalk correction circuit 19) that corrects crosstalk generated between the respective lines using the reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the crosstalk amount. Will be described with reference to the flowchart of FIG.

  First, the number of read lines is set to the number of sub-scanning lines of the document 9 in accordance with the size of the document 9 to be read (step b1).

  Next, the movement of the line of sight of the color CCD 7 is started (step b2), and reading is started when the three lines of sight of the color CCD 7 reach the leading edge of the document 9 for the first time (step b3).

  Simultaneously with the start of reading, the crosstalk correction circuit 19 of the image processing LSI 13 performs the crosstalk correction in real time (step b4).

  Next, delay adjustment is performed by the inter-line correction circuit 20 using the image data from which crosstalk has been removed, and the gap between lines is corrected (step b5). Gap correction by the interline correction circuit 20 is already known.

  By repeatedly performing the processing from step b2 to step b5 up to the set number of sub-scanning lines of the original (step b6), final image data subjected to crosstalk correction and gap correction is obtained.

  Here, the operation of the crosstalk correction circuit 19 functioning as a crosstalk correction output means (the process of step b4) will be described in detail.

  First, let the gain amount obtained by the gain control circuit 16 be R (red) / G (green) / B (blue) respectively αr, αg, αb. Also, the image data of each line before crosstalk correction read at the time of scanning is R2, G2, and B2, respectively.

  The reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the crosstalk amount for each line are stored in the crosstalk correction memory 14, and the reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb, gain values αr, αg, αb, read image data R2, G2, B2 before crosstalk correction, and resolution of the A / D conversion circuit 17, that is, a maximum value P of density expression (for example, 8 In the case of bit output, 256) is used to obtain corrected data R3, G3, B3 by the following equations (7) to (9).

R3 = R2- (ΔGr · αg) (R2-G2) / P- (ΔBr · αb) (R2-B2) / P (7)
In other words, the reference value ΔGr of the crosstalk amount is the G (green) read by positioning each line of R (red) / G (green) / B (blue) on the white reference plane with the gain = 1. Deviation between the image data G1 of the line and the image data Gr of the G (green) line when only the R (red) line is positioned in the black line portion and the G (green) line is positioned on the white reference plane Gr-G1, in short, is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of density expression, so the density deviation R2-G2 of the actually scanned image data Is multiplied by the gain amount αg of the G (green) line at the time of scanning and the maximum value P of the density expression, and αg (R2-G2) / P is weighted to the reference value ΔGr of the crosstalk amount As a result, the influence of crosstalk corresponding to the difference in density and reflectance of the document 9 or the deterioration of the light source is required. Similarly, the reference value ΔBr of the crosstalk amount is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of density expression, and thus is actually scanned. The value obtained by multiplying the density deviation R2-B2 of the obtained image data by the gain amount αb of the B (blue) line at the time of scanning and the maximum value P of the density expression represents the crosstalk amount reference value ΔBr with αb ( By performing the weighting of (R2-B2) / P, the influence of crosstalk corresponding to the difference in density and reflectance of the document 9 or the deterioration of the light source is required. Therefore, finally, a subtraction process R2 for removing the influence of (ΔGr · αg) (R2-G2) / P and (ΔBr · αb) (R2-B2) / P from the image data R2 of the R (red) line. By performing-(ΔGr · αg) (R2-G2) / P- (ΔBr · αb) (R2-B2) / P, R can be used regardless of differences in the density or reflectance of the document 9 or deterioration of the light source. Appropriate image data R3 can be obtained as the image data of the (red) line.

G3 = G2- (ΔRg · αr) (G2-R2) / P- (ΔBg · αb) (G2-B2) / P (8)
In this case as well, as described above, the reference value ΔRg of the crosstalk amount is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of density expression, and thus is actually scanned. The crosstalk amount reference value ΔRg is set to αr () by a proportional relationship between the density deviation G2-R2 of the obtained image data multiplied by the gain amount αr of the R (red) line at the time of scanning and the maximum value P of the density expression. By performing the weighting of (G2-R2) / P, the influence of crosstalk corresponding to the difference in the density and reflectance of the document 9 or the deterioration of the light source is required. In addition, since the reference value ΔBg of the crosstalk amount is also a value obtained corresponding to the gradation difference P between white and black, which is the maximum density expression, the density of the actually scanned image data By multiplying the deviation G2-B2 by the gain amount αb of the B (blue) line at the time of scanning and the maximum value P of the density expression, αb (G2-B2) / By performing the weighting of P, the influence of crosstalk corresponding to the difference in density and reflectance of the document 9 or the deterioration of the light source is required. Therefore, finally, the subtraction process G2 for removing the influence of (ΔRg · αr) (G2-R2) / P and (ΔBg · αb) (G2-B2) / P from the image data G2 of the G (green) line. By performing-(ΔRg · αr) (G2-R2) / P- (ΔBg · αb) (G2-B2) / P, regardless of the difference in the density or reflectance of the document 9 or the deterioration of the light source, the G Appropriate image data G3 can be obtained as the image data of the (green) line.

B3 = B2- (ΔRb · αr) (B2-R2) / P- (ΔGb · αg) (B2-G2) / P (9)
Also in this case, since the reference value ΔRb of the crosstalk amount is a value obtained corresponding to the gradation difference P between white and black, which is the maximum density expression value, the actually scanned image data The density difference B2-R2 is multiplied by the gain amount αr of the R (red) line at the time of scanning and the maximum value P of the density expression is proportional to the crosstalk amount reference value ΔRb by αr (B2-R2 ) / P weighting can be used to determine the influence of crosstalk in accordance with the density of the document 9 and the difference in reflectance or the deterioration of the light source. In addition, since the reference value ΔGb of the crosstalk amount is also a value obtained corresponding to the gradation difference P between white and black, which is the maximum density expression, the density of the actually scanned image data Based on the proportional relationship between the deviation G2-B2 multiplied by the gain amount αg of the G (green) line at the time of scanning and the maximum value P of the density expression, αg (B2-G2) / P Thus, the influence of crosstalk corresponding to the difference in density and reflectance of the document 9 or the deterioration of the light source is required. Therefore, finally, the subtraction process B2 for removing the influence of (ΔRb · αr) (B2-R2) / P and (ΔGb · αg) (B2-G2) / P from the image data B2 of the B (blue) line. By performing-(ΔRb · αr) (B2-R2) / P- (ΔGb · αg) (B2-G2) / P, the difference between the density of the document 9 and the reflectance, the deterioration of the light source, etc. Appropriate image data B3 can be obtained as image data of the (blue) line.

  Then, the image data R3, G3, and B3 after crosstalk correction obtained in this way are stored in the interline correction memory 15 as described above, and the delay between each line (for example, 8 lines) is taken into consideration. As in the prior art, the inter-line correction circuit 20 corrects the gap between the lines, and finally obtains appropriate image data.

  FIG. 6 is a conceptual diagram showing an outline of the configuration of the mechanical part of a scanner crosstalk correction device 1 ′ according to another embodiment corresponding to a high-speed scanner.

  In the case of the above-described scanner crosstalk correction device 1, the color CCD 7 runs under the white reference plane 4 in order for the color CCD 7 to read the black line portion 5 between the white reference plane 4 and the white reference plane 6. In the meantime, the color CCD 7 needs to reach the target speed, and if this is to be applied to a high-speed scanner, the width of the white reference surface 4 with respect to the sub-scanning direction must be increased. Detrimental effects such as enlargement may occur.

  The scanner crosstalk correction apparatus 1 ′ shown in FIG. 6 is for solving such a problem and dealing with the speeding up of the scanner without increasing the size of the surface for reading the white reference and the black reference. .

  As shown in FIG. 6, the main part of the mechanism of the scanner crosstalk correction device 1 ′ is a black reference surface (black reference surface) placed on the contact glass 3 for placing the document 9 on the scanner 2 ′. ) 22, 24, a white line portion (white reference) 23 provided on the same plane so as to be sandwiched between the black reference surfaces 22, 24, and a white reference corresponding to the white reference surface 4 of the above-described embodiment. It is constituted by the surface 21.

  The width of the black reference planes 22 and 24 in the sub-scanning direction is more than twice the line interval of the color CCD 7 and has a sufficient width in the main scanning direction. Further, the white line portion 23 and the white reference surface 21 have the same whiteness as the white reference surfaces 4 and 6 in the above-described embodiment, and similarly to the black line portion 5 in the above-described embodiment, the width in the sub-scanning direction is It is formed narrower than the width between the lines of the color CCD 7. Therefore, for example, if the color CCD 7 is constituted by a CCD having an interval of 8 lines and the resolution in the sub-scanning direction of the scanner 2 ′ is 600 dpi, the width of the white line portion 23 is 25.4 mm / 600 × (8-1). ) ≈0.3mm, and it is necessary to make the width less than 0.3mm. Since the black reference surface 22 and the black reference surface 24 have substantially the same characteristics, it can be said that the white line portion 23 is provided in the black reference surface which is entirely constituted by the black reference surfaces 22 and 24.

  The configuration of the color CCD 7 and the lamp 8 mounted on the scanner 2 'side is the same as that in the above-described embodiment.

  The control unit of the scanner crosstalk correction apparatus 1 ′ can use the control unit 10 shown in FIG. 2 as it is. In this embodiment, the processing operation of the crosstalk amount calculation means in the crosstalk correction circuit 19 is performed. Different from the previous embodiment.

  Next, the processing of the crosstalk amount calculating means (crosstalk correction circuit 19) of this embodiment will be described with reference to the flowchart of FIG.

  First, the lamp 8 is turned on (step c1), the line of sight of the color CCD 7 is moved, and stopped just below the white reference plane 21 (step c2, step c3). The stop position is in the vicinity of the center of the white reference surface 21 so that the three lines of the color CCD 7 can be read stably and the white of the white reference surface 21 can be read.

  Then, a weight is applied until the light quantity of the lamp 8 is stabilized, and the color CCD 7 reads the white of the white reference surface 21 (step c4).

  Next, a desired gain value is set for the gain control circuit 16 of the analog front end unit (AFE) 12, and when black correction is necessary, a correction value BLK1 for black correction is calculated (step c5). At this time, the correction value BLK1 set for the black correction circuit 18 is the value of the image data when the black reference plane 22 is read in consideration of the crosstalk amount of the color CCD 7 estimated in advance, that is, black. An offset is applied so that the values of the reference image data R1, G1, B1 do not become the limit value 0 (jet black) of density expression. Specifically, the correction value BLK1 may be set so that the values of the black reference image data R1, G1, B1 including the influence of crosstalk do not fall below the lower limit value 0 of the density expression. Then, the correct black correction value BLK1 is stored in the crosstalk correction memory 14. If necessary, a shading correction calculation is performed to compensate for the non-uniformity of the light amount of the lamp 8.

  Next, in order to read the black reference surface 22, the line of sight of the color CCD 7 is moved to the left in FIG. 6 along the sub-scanning direction (step c6, step c7), and the read image data of the black reference surface 22 is obtained. The black reference image data R1, G1, B1 is stored in the crosstalk correction memory 14 (step c8).

Here, in order to obtain a run-up distance for accelerating the moving speed of the color CCD 7 to a target speed corresponding to the reading speed of the scanner 2 ′, the line of sight of the color CCD 7 is once aligned along the sub-scanning direction in the right side of FIG. It is moved in the direction (step c9, step c10).
The moving distance at this time is determined by the reading speed of the scanner 2 ′, and is a distance at which the reading speed is stabilized in the area within the white reference plane 21.

  Next, the timing to start reading and the timing to end reading are set along the sub-scanning direction (step c11). The starting position of the reading is that the run of the color CCD 7 is completed and the line of the line that is delayed most in the line of the color CCD 7, that is, in this embodiment, the B (blue) line is changed from the white reference surface 21 to the black reference surface. 22 is the moment of transition. Similarly, the reading end position is the moment when the eye line of the most delayed line shifts from the white line portion 23 to the black reference plane 24. Specifically, the transition from the white area to the black area can be confirmed by detecting the falling edge of the signal output from the B (blue) line.

  Next, in order to move the line of sight of the color CCD 7 to the left, a run is started, and reading of an image by the color CCD 7 is executed (step c12 to step c14). Since the reading start position and reading end position are set, the image data that is actually read and stored in the inter-line correction memory 15 has a B (blue) line from the white reference plane 21 to the black reference plane. It is limited to the time from when the transition to 22 is reached to when the B (blue) line transitions from the white line portion 23 to the black reference plane 24, and the image data of the previous run section is not saved. The image data read here is stored in the inter-line correction memory 15 as image data when each line reads the white line portion 23 by the crosstalk correction circuit 19 of the image processing LSI 13.

  Thereafter, the movement of the line of sight of the color CCD 7 is stopped, and the reference value of the crosstalk amount is calculated (step c15).

  A method for calculating the reference value of the crosstalk amount will be described in detail below.

  First, the image data when the white line portion 23 is read from the inter-line correction memory 15 of the image processing LSI 13 is read.

  Then, for each read R (red) / G (green) / B (blue) data, the average value R0 in the range from the reading range PIXS to PIXE in the main scanning direction as in the above-described embodiment. , G0, B0 are calculated and stored in the inter-line correction memory 15. It is assumed that the reading range PIXS to PIXE is selected near the center in the main scanning direction where the light amount of the lamp 8 is brightest. This average value calculation process is repeated for the number of sub-scanning lines read from the white line portion 23.

  When the contents of the average values R0, G0, B0 stored in the interline correction memory 15 are viewed in time series, a waveform as shown in FIG. 8 is obtained.

  Here, an 8-line color CCD 7 is taken as an example, and the width of the white line portion 23 is described as 7 lines.

  First, an average value for seven lines is obtained for the pixel levels of the remaining G (green) line and B (blue) line when the R (red) line reads the white line portion 23 and the value changes.

  The obtained average values are set as black image data Gr and Br, respectively.

Here, the image data (see step c8) obtained by reading the black reference surface 22 in advance is set as the black reference image data R1, G1, B1, and the reference value of the crosstalk amount from the R (red) line to the G (green) line. Is ΔGr, and the reference value of the crosstalk amount from the R (red) line with respect to the B (blue) line is ΔBr, the reference values ΔGr and ΔBr of the crosstalk amount are expressed by the following equations (10) and (11). can get.
ΔGr = G1-Gr (10)
ΔBr = B1-Br (11)

Similarly, the average values of seven lines of the R (red) line and the B (blue) line when the G (green) line is changed by reading the white line portion 23 are set as black image data Rg and Bg, respectively. At this time, if the reference value of the crosstalk amount from the G (green) line for the R (red) line is ΔRg and the reference value of the crosstalk amount from the G (green) line for the B (blue) line is ΔBg, The reference values ΔRg and ΔBg of the talk amount are obtained by the following equations (12) and (13).
ΔRg = R1-Rg (12)
ΔBg = B1-Bg (13)

Similarly, the average values for the seven lines of the R (red) line and the G (green) line when the B (blue) line is changed by reading the white line portion 23 are respectively determined as white image data Rb, Gb. To do. At this time, if the reference value of the crosstalk amount from the B (blue) line to the R (red) line is ΔRb and the crosstalk amount from the B (blue) line to the G (green) line is ΔGb, the crosstalk amount is The reference values ΔRb and ΔGb are obtained by the following equations (14) and (15).
ΔRb = R1-Rb (14)
ΔGb = G1-Gb (15)

  After calculating the reference values of the crosstalk amounts in the above formulas (10) to (15), the crosstalk amount reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb for the obtained lines are crosstalk corrected. The correct black correction value BLK1 stored in the crosstalk correction memory 14 is read and set in the black correction circuit 18 of the analog front end unit (AFE) 12 at the end. (Step c17), the calculation process related to the reference value of the crosstalk amount is terminated.

Expressions (10) to (15) in this embodiment correspond to Expressions (1) to (6) of the above-described embodiment, but in the above-described embodiment, the white reference image data R1, G1, B1 The reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the crosstalk amount are obtained with reference to the black reference image data R1, G1, B1 in the present embodiment. Since ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb are obtained, the signs are reversed as compared with the above-described embodiment.
It is also possible to invert the sign in the crosstalk correction process described later. In such a case, the above-described equations (1) to (6) are applied as they are and the crosstalk is applied. The reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the quantities may be obtained.

  Regarding the processing of the crosstalk correction output means (crosstalk correction circuit 19) for correcting the crosstalk generated between the respective lines using the reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the crosstalk amount, Since the overall processing operation is generally the same as that of the above-described embodiment, description will be made based on the flowchart of FIG.

  First, the number of read lines is set to the number of sub-scanning lines of the document 9 in accordance with the size of the document 9 to be read (step b1).

  Next, the movement of the line of sight of the color CCD 7 is started (step b2), and reading is started when the three lines of sight of the color CCD 7 reach the leading edge of the document 9 for the first time (step b3).

  Simultaneously with the start of reading, the crosstalk correction circuit 19 of the image processing LSI 13 performs the crosstalk correction in real time (step b4).

  Next, delay adjustment is performed by the inter-line correction circuit 20 using the image data from which crosstalk has been removed, and the gap between lines is corrected (step b5). Gap correction by the interline correction circuit 20 is already known.

  By repeatedly performing the processing from step b2 to step b5 up to the set number of sub-scanning lines of the original (step b6), final image data subjected to crosstalk correction and gap correction is obtained.

  Here, the operation of the crosstalk correction circuit 19 functioning as a crosstalk correction output means (the process of step b4) will be described in detail. In the present embodiment, the processing for crosstalk correction is slightly different from the above-described embodiment.

  First, the image data of each line before crosstalk correction read at the time of scanning is set as R2, G2, and B2, respectively. However, these image data are different from R2, G2, and B2 in the above-described embodiment, and are values that have already undergone gain adjustment.

  The reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb of the crosstalk amount for each line are stored in the crosstalk correction memory 14, and the reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb, the read image data R2, G2, B2 before crosstalk correction, and the resolution of the A / D conversion circuit 17, that is, the maximum value P of the density expression (for example, 256 in the case of 8-bit output) are used. Thus, the corrected data R3, G3, B3 are obtained by the following equations (16) to (18).

R3 = R2-ΔGr (R2-G2) / P-ΔBr (R2-B2) / P (16)
In other words, the reference value ΔGr of the crosstalk amount is the image data G1 of the G (green) line read by positioning each R (red) / G (green) / B (blue) line on the black reference plane. Deviation G1-Gr from the image data Gr of the G (green) line when only the R (red) line is positioned in the white line portion and the G (green) line is positioned on the black reference plane. Since it is the value obtained corresponding to the gradation difference P between white and black, which is the maximum value of expression, the value of density deviation R2-G2 of the actually scanned image data and the maximum of density expression Due to the proportional relationship with the value P, the reference value ΔGr of the crosstalk amount is weighted by (R2-G2) / P, so that the influence of the crosstalk according to the difference in the density of the document 9, the reflectance, the deterioration of the light source, etc. Is required. Similarly, the reference value ΔBr of the crosstalk amount is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of density expression, and thus is actually scanned. The density of the document 9 is obtained by weighting (R2-B2) / P to the reference value ΔBr of the crosstalk amount according to the proportional relationship between the density deviation R2-B2 of the measured image data and the maximum value P of the density expression. And the influence of crosstalk according to the difference in reflectivity or the deterioration of the light source is required. Therefore, finally, subtraction processing R2-ΔGr (R2-G2) / which removes the influence of ΔGr (R2-G2) / P and ΔBr (R2-B2) / P from the image data R2 of the R (red) line. By performing P-ΔBr (R2-B2) / P, appropriate image data R3 is obtained as image data of the R (red) line regardless of the density or reflectance of the document 9 or the deterioration of the light source. be able to.

G3 = G2-ΔRg (G2-R2) / P-ΔBg (G2-B2) / P (17)
In this case as well, as described above, the reference value ΔRg of the crosstalk amount is a value obtained corresponding to the gradation difference P between white and black, which is the maximum value of density expression, and thus is actually scanned. The density of the document 9 is obtained by weighting the reference value ΔRg of the crosstalk amount by (G2-R2) / P according to the proportional relationship between the density deviation G2-R2 of the image data and the maximum value P of the density expression. And the influence of crosstalk according to the difference in reflectivity or the deterioration of the light source is required. In addition, since the reference value ΔBg of the crosstalk amount is also a value obtained corresponding to the gradation difference P between white and black, which is the maximum density expression, the density of the actually scanned image data The difference in the density and reflectance of the document 9 is obtained by weighting (G2-B2) / P to the reference value ΔBg of the crosstalk amount according to the proportional relationship between the deviation G2-B2 value and the maximum value P of the density expression. Alternatively, the influence of crosstalk corresponding to deterioration of the light source or the like is required. Therefore, finally, a subtraction process G2-ΔRg (G2-R2) / that removes the influence of ΔRg (G2-R2) / P and ΔBg (G2-B2) / P from the image data G2 of the G (green) line. By performing P-ΔBg (G2-B2) / P, appropriate image data G3 is obtained as image data of the G (green) line regardless of the density or reflectance of the document 9 or the deterioration of the light source. be able to.

B3 = B2-ΔRb (B2-R2) / P-ΔGb (B2-G2) / P (18)
Also in this case, since the reference value ΔRb of the crosstalk amount is a value obtained corresponding to the gradation difference P between white and black, which is the maximum density expression value, the actually scanned image data The density and reflectance of the document 9 can be obtained by weighting the reference value ΔRb of the crosstalk amount by (B2-R2) / P according to the proportional relationship between the density deviation B2-R2 and the maximum density expression P. Therefore, the influence of crosstalk is required according to the difference of the light source or the deterioration of the light source. In addition, since the reference value ΔGb of the crosstalk amount is also a value obtained corresponding to the gradation difference P between white and black, which is the maximum density expression, the density of the actually scanned image data Depending on the proportional relationship between the value of deviation G2-B2 and the maximum value P of the density expression, weight difference (B2-G2) / P is applied to the reference value ΔGb of the crosstalk amount, or the difference in density and reflectance of the document 9 or The influence of crosstalk corresponding to deterioration of the light source is required. Therefore, finally, a subtraction process B2-ΔRb (B2-R2) / which removes the influence of ΔRb (B2-R2) / P and ΔGb (B2-G2) / P from the image data B2 of the B (blue) line. By performing P-ΔGb (B2-G2) / P, proper image data B3 is obtained as the image data of the B (blue) line regardless of the density or reflectance of the document 9 or the deterioration of the light source. be able to.

  The processing of Expressions (16) to (18) related to crosstalk correction is basically the same as Expressions (7) to (9) of the above-described embodiment, but in this embodiment, scanning is performed. Since the image data R2, G2, and B2 of each line read at this time are values after gain adjustment, the gains (specifically, αr, αg, αb) as shown in the equations (7) to (9) The process using is unnecessary.

  Then, the image data R3, G3, and B3 after crosstalk correction obtained in this way are stored in the interline correction memory 15 as described above, and the delay between each line (for example, 8 lines) is taken into consideration. As in the prior art, the inter-line correction circuit 20 corrects the gap between the lines, and finally obtains appropriate image data.

  As described above, in each of the above-described embodiments, the black line portion 5 as shown in FIG. 1 and the white line portion 23 as shown in FIG. 6 are added to the conventional scanner 2 and scanner 2 ′. In the pre-processing stage of scanning the original, these white reference and black reference are read, and the reference values ΔGr, ΔBr of the crosstalk amount for each line of R (red) / G (green) / B (blue), ΔRg, ΔBg, ΔRb, ΔGb are quantitatively obtained in advance and stored in the crosstalk correction memory 14. Further, when scanning a document, the image data R2, G2, B2 of each line read at the time of scanning are , Deviations between the image data R2, G2, B2 and the image data (G2, B2), (R2, B2), (R2, G2) of the other lines except for the above, that is, R (red ) Line for (R2-G2), (R2-B2), G (green) line for (G2-R2), (G2-B2), B (blue) line Of (B2-R2) and (B2-G2) Each deviation (R2-G2), (R2-B2), (G2-R2), (G2-B2), (B2-R2), (B2-G2) is the maximum value p of the concentration expression. (For example, 256) and the crosstalk amount reference values ΔGr, ΔBr, ΔRg, ΔBg, ΔRb, ΔGb corresponding to the other lines, that is, ΔGr for the R (red) line , ΔBr, G (green) lines for ΔRg, ΔBg, B (blue) lines based on the values of ΔRb, ΔGb, R (red) / G (green) / B (blue) Since the image data R2, G2, and B2 of each line are corrected in real time to eliminate crosstalk, the scanning conditions change due to differences in the density and reflectance of the document 9 or deterioration of the light source. It is possible to cope dynamically and suppress crosstalk between lines generated in a general-purpose three-line CCD accurately and inexpensively.

In particular, when a color CCD having a relatively large line interval such as 8 lines is used, crosstalk between R (red) and green G (green) is noticeably generated with respect to B (blue) having low sensitivity. By applying these embodiments, it is possible to effectively reduce the influence of crosstalk.
As a result, since the proportion affected by the line spacing is reduced, the range of choices for color CCDs is widened, which is effective in improving the degree of design freedom and reducing the manufacturing costs of the scanners 2 and 2 ′. .

  Further, since at least a part of the data required for crosstalk correction is stored using the inter-line correction memory 15, the size of the memory required for crosstalk correction (for crosstalk correction). The storage capacity of the memory 14) can be greatly reduced.

It is the conceptual diagram which showed the outline of the structure of the mechanism part of the crosstalk correction apparatus for scanners of one Embodiment to which the crosstalk correction method of this invention was applied. It is the functional block diagram which showed the outline of the structure of the control part of the crosstalk correction apparatus for the scanner. It is the flowchart which showed the outline of the process in connection with calculation of the reference value of the amount of crosstalk. It is the conceptual diagram which showed the time-sequential change of the data of each line of R (red) / G (green) / B (blue) preserve | saved at the memory for correction | amendment between lines. It is the flowchart which showed the outline of the process in connection with crosstalk correction. It is the conceptual diagram which showed the outline of the structure of the mechanism part of the crosstalk correction apparatus for scanners of other one Embodiment. It is the flowchart which showed the outline of the process in connection with calculation of the reference value of the amount of crosstalk. It is the conceptual diagram which showed the time-sequential change of the data of each line of R (red) / G (green) / B (blue) preserve | saved at the memory for correction | amendment between lines.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Crosstalk correction device for scanners 1 'Crosstalk correction device for scanners 2 Scanner 2' Scanner 3 Contact glass 4 White reference plane (white reference)
5 Black line (black reference)
6 White reference plane (white reference)
7 Color CCD
8 Lamp 9 Document 10 Control unit 11 Image input unit 12 Analog front end unit 13 Image processing LSI
14 Crosstalk correction memory 15 Line-to-line correction memory 16 Gain control circuit 17 A / D conversion circuit 18 Black correction circuit 19 Crosstalk correction circuit (crosstalk amount calculation means, crosstalk correction output means)
20 Interline correction circuit 21 White reference plane 22 Black reference plane (black reference)
23 White line part (white standard)
24 Black reference plane (black reference)

Claims (10)

A scanner crosstalk correction device for correcting crosstalk generated between lines when a document is read with a color CCD having different light receiving characteristics provided with a certain line interval,
Crosstalk amount calculating means for reading a white reference and a black reference by the color CCD at a pre-processing stage for scanning a document and obtaining a reference value of a crosstalk amount that one line gives to each other line for each line; ,
A cross that applies correction for eliminating crosstalk in real time to the image data of each line based on the image data of each line scanned and read from the original and the reference value of the crosstalk amount corresponding to each other line. A crosstalk correction device for a scanner, comprising: a talk correction output unit. A white reference composed of a white reference surface disposed on a contact glass for placing an original, and a black reference composed of a black line portion provided in the white reference surface,
The crosstalk amount calculation means positions the color CCD of each line on the white reference plane and reads the white reference image data of each line and one of the lines positioned on the black line portion. Based on the deviation between the white image data of each other line read and each white reference image data of the other line, the reference value of the crosstalk amount given to each other line by the one line is set to each line. Ask for every
The crosstalk correction output means is configured such that each deviation between the image data of each line scanned and read from the document and the image data and the image data of each of the other lines other than this is the maximum value of the density expression. It is configured to apply correction for eliminating crosstalk to the image data of each line on the basis of the ratio to the other line and the reference value of the crosstalk amount corresponding to each of the other lines. The scanner crosstalk correction device according to claim 1.   The scanner crosstalk correction device according to claim 2, wherein the black line portion is formed to be narrower than a width between lines of the color CCD.   The crosstalk amount calculating means temporarily stores the white image data in an interline correction memory of an interline correction circuit used for delay adjustment of image data for each line, and uses the average value of each white image data to cross 4. The crosstalk correction device for a scanner according to claim 2, wherein a reference value of the talk amount is obtained. A black reference composed of a black reference surface disposed on a contact glass for placing an original, and a white reference composed of a white line portion provided in the black reference surface,
The crosstalk amount calculating means positions the color CCD of each line on the black reference plane and reads the black reference image data of each line, and positions one of the lines on the white line portion. Based on the deviation between the read black image data of each other line and each black reference image data of the other line, the reference value of the crosstalk amount given to each other line by the one line is set for each line. Seeking
The crosstalk correction output means is configured such that each deviation between the image data of each line scanned and read from the document and the image data and the image data of each of the other lines other than this is the maximum value of the density expression. It is configured to apply correction for eliminating crosstalk to the image data of each line on the basis of the ratio to the other line and the reference value of the crosstalk amount corresponding to each of the other lines. The scanner crosstalk correction device according to claim 1.   6. The scanner crosstalk correction device according to claim 5, wherein the white line portion is formed narrower than a width between lines of the color CCD.   7. The crosstalk amount calculating means is configured to offset black correction so that the black reference image data does not become a value near the limit of density expression. Crosstalk correction device for scanners.   The crosstalk amount calculation means temporarily stores the black image data in an interline correction memory of an interline correction circuit used for delay adjustment of image data for each line, and uses the average value of each black image data to cross 8. The scanner crosstalk correction device according to claim 5, wherein a reference value of the talk amount is obtained. A crosstalk correction method for correcting crosstalk that occurs between lines when a document is read with a color CCD having different light receiving characteristics provided with a certain line interval,
In the pre-processing stage of scanning the document, the white CCD image data of each line read with the color CCD of each line positioned on the white reference and one of the lines positioned on the black reference are read. On the basis of the deviation between the white image data of each other line and the respective white reference image data of the other line, the reference value of the crosstalk amount given to each of the other lines by the one line is determined for each line. Asking
When scanning an original, the ratio of each deviation between the image data of each line read in the scan and the image data and the image data of each of the other lines excluding this to the maximum value of the density expression, A crosstalk correction method, wherein correction for eliminating crosstalk is performed in real time on the image data of each line based on the reference value of the crosstalk amount corresponding to each of the other lines. A crosstalk correction method for correcting crosstalk that occurs between lines when a document is read with a color CCD having different light receiving characteristics provided with a certain line interval,
In the pre-processing stage of scanning the original, the black reference image data of each line read with the color CCD of each line positioned at the black reference and one of the lines positioned at the white reference and read On the basis of the deviation between the black image data of each other line and each black reference image data of the other line, the reference value of the crosstalk amount given to each other line by the one line is determined for each line. Asking
When scanning an original, the ratio of each deviation between the image data of each line read in the scan and the image data and the image data of each of the other lines excluding this to the maximum value of the density expression, A crosstalk correction method, wherein correction for eliminating crosstalk is performed in real time on the image data of each line based on the reference value of the crosstalk amount corresponding to each of the other lines.

Images