JP2018011157A - Read image correction device, read image correction method, and computer program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To correct an image that is read by a one-line scanner, more appropriately than the prior arts.SOLUTION: For a red target pixel subjected to correction in a read image, a feature amount representing a feature of a change from surrounding red pixels is calculated (Figures 15(A) and (C)). Gradations of other colors (that are green and blue) in the target pixel are calculated based on gradations of surrounding green pixels and gradations of blue pixels (Figure 15(B)). The gradation of the pixel and the gradation of the blue pixel are then corrected based on the feature amount (Figure 15(D)). For a blue target pixel and a green target pixel, feature amounts are also similarly calculated, gradations of the other colors are calculated, and the gradations are corrected based on the feature amounts.SELECTED DRAWING: Figure 15

Description

  The present invention relates to a technique for correcting an image read by a line scanner.

  Conventionally, so-called one-line scanners have become widespread. The one-line scanner has light sources of RGB (Red Green Blue) colors, and when one line is read, the light source of one color of RGB is turned on, and the colors that are turned on for each line are sequentially turned on. It is a line scanner to switch. It may be called “one line sensor” or “one line scan camera”.

  When a color image is scanned by a one-line scanner, if the image is an image whose gradation gradually changes linearly in the sub-scanning direction, for example, to match the red phase to the green phase, interpolation is performed. What is necessary is just to calculate the gradation of red. Specifically, the red gradation of a specific pixel may be calculated based on the gradations of a plurality of red pixels around the green specific pixel (Patent Document 1).

  However, when reading black characters on white paper, the gradation is adjusted in the conventional way to match the phase of one color (eg, green) with the other color (eg, red and blue). May cause a difference even though the gray levels of red, blue, and green should originally be equal. In particular, a difference is likely to occur near the edge of a character.

  Although it is originally a black character, the larger the difference is, the easier it is to be mistakenly recognized as a color character.

  Therefore, the following techniques have been proposed. When the selection to perform ACS (Auto Color Select) is made, the sub-scanning resolution is increased (Patent Document 2).

  The image reading device is lit in one color during the reading time of one line, and a light source unit that sequentially switches the color of the light source that is turned on for each line, and reading for one color per line based on the reflected light of the document. An image sensor to perform, a data generation unit that generates image data of a light source color that is a color of a light source that is turned on at the time of reading, a memory that stores image data generated by the data generation unit for a plurality of lines, and stored in the memory A residual color component generation unit that generates pixel values of color components other than the light source color based on pixel values of peripheral pixels of the target pixel using image data for a plurality of lines (Patent Document 3).

JP 2002-247292 A JP 2013-85172 A JP 2014-27508 A

  According to the technique described in Patent Document 2, the speed of reading an image is slow. According to the technique described in Patent Document 3, it is possible to prevent the pixel at the edge of the black character from being erroneously recognized as a color pixel. However, there is a possibility that a color pixel is erroneously recognized as a black pixel.

  In view of such problems, an object of the present invention is to enable an image read by a one-line scanner to be corrected more suitably than in the past.

  A read image correction apparatus according to an aspect of the present invention is a read image correction apparatus that corrects a read image that is an image read by a one-line scanner while sequentially turning on light sources of a plurality of colors for each line. A feature amount representing a characteristic of a change in the target pixel as a correction target in the read image from the periphery of the target pixel is represented by the gradation of the target pixel obtained by the one-line scanner. When the reading light source used for reading the target pixel among the light sources of each of the plurality of colors in the read image is turned on by the one line scanner A feature amount calculation means for calculating by a second reading gradation that is a gradation of a pixel that is read and is adjacent to the target pixel in the sub-scanning direction; A non-read color gradation that is a gradation of a non-read color that is a color different from the color corresponding to the reading light source is the non-read color of the light sources of the plurality of colors in the read image. Non-reading color gradation calculation that is calculated based on the third reading gradation of each of a plurality of pixels that are read by the one-line scanner and close to the target pixel in the sub-scanning direction when the light source is turned on And correction means for calculating a correction gradation by correcting the non-read color gradation based on the feature amount.

  Preferably, the feature amount calculation unit calculates the feature amount by adding a value obtained by applying a second-order differential filter to the first reading gradation and the second reading gradation.

  According to the present invention, an image read by a one-line scanner can be corrected more suitably than in the past.

1 is a diagram illustrating an example of an appearance of an image forming apparatus. 2 is a diagram illustrating an example of a hardware configuration of an image forming apparatus. FIG. It is a figure which shows the example of a structure of a scan unit. 2 is a diagram illustrating an example of a functional configuration of an image forming apparatus. FIG. It is a figure which shows the example of image data. It is a figure which shows the example of an image. It is a figure which shows the example of the positional relationship of the pixel of reading color, and the pixel of non-reading color. It is a figure which shows the example of the change of the gradation of the continuous pixel with the same coordinate of a subscanning direction. It is a figure for demonstrating the example of the method of calculation of a reference gradation. It is a figure which shows the example of the reference gradation of the non-reading color of each pixel. It is a figure which shows the example of the positional relationship of the continuous pixel with the same reading color. It is a figure which shows the example of the method of application of a secondary differential filter. It is a figure which shows the example of the change feature-value of each pixel. It is a figure which shows the example of the correction | amendment gradation of the non-reading color of each pixel. It is a figure for demonstrating the specific example of calculation of a change feature-value, and correction | amendment of a reference gradation. It is a flowchart explaining the example of the flow of a 1st gradation determination process. It is a flowchart explaining the example of the flow of a 2nd gradation determination process. It is a figure which shows the example of the result of the process in the black character reading mode by a gradation determination part. It is a figure which shows the example of the result of the process in the color image reading mode by a gradation determination part. It is a flowchart explaining the example of the flow of the whole process of the process by a read image correction program. It is a flowchart explaining the example of the flow of a reference | standard gradation correction process. It is a figure which shows the example of the positional relationship of an attention pixel and the pixel before and behind it. It is a figure which shows the example of the method of application of a secondary differential filter.

  FIG. 1 is a diagram illustrating an example of the appearance of the image forming apparatus 1. FIG. 2 is a diagram illustrating an example of a hardware configuration of the image forming apparatus 1. FIG. 3 is a diagram illustrating an example of the configuration of the scan unit 10e.

  An image forming apparatus 1 shown in FIG. 1 is an apparatus generally called a multi function peripheral or MFP (Multi Function Peripherals), and is an apparatus that integrates functions such as copying, scanning, network printing (PC printing), and fax. .

  The image forming apparatus 1 can exchange image data with an apparatus such as a personal computer or a fax terminal via a communication line such as a LAN (Local Area Network), a public line, or the Internet.

  As shown in FIG. 1 or 2, the image forming apparatus 1 includes a system controller 10, an auxiliary storage device 10d, a scan unit 10e, a print unit 10f, a NIC (Network Interface Card) 10g, a modem 10h, a touch panel display 10i, and an operation key panel. 10j, an input image processing unit 10k, an output image processing unit 10m, and the like.

  The system controller 10 includes a CPU (Central Processing Unit) 10a, a RAM (Random Access Memory) 10b, a ROM (Read Only Memory) 10c, and the like, and executes jobs such as copying, scanning, e-mail transmission, and network printing. Therefore, overall control of the image forming apparatus 1 is performed.

  The ROM 10c stores a program for executing each job described above. Note that all or part of these programs may be stored in the auxiliary storage device 10d.

  Furthermore, a read image correction program 10P (see FIG. 4) is stored in the ROM 10c or the auxiliary storage device 10d. The read image correction program 10P is appropriately loaded into the RAM 10b and executed by the CPU 10a. According to the read image correction program 10P, color blur that is likely to occur in a one-line scanner can be reduced as compared with the conventional case. Details of the read image correction program 10P will be described later.

  A hard disk or SSD (Solid State Drive) is used as the auxiliary storage device 10d.

  These programs are appropriately loaded into the RAM 10b and executed by the CPU 10a. Further, the RAM 10b appropriately stores various data (for example, setting data) when executing these programs, data (for example, image data) obtained by the execution, and the like.

  The scan unit 10e reads an image such as a photograph, a character, a picture, a chart, or the like written on the paper of the document and generates image data.

  Specifically, as shown in FIG. 3, the scan unit 10e includes a scanner 10e1, a reading slit 10e2, a platen glass 10e3, an ADF (Auto Document Feeder) 10e4, and the like.

  The ADF 10e4 conveys the set sheets one by one to the reading slit 10e2. The scanner 10e1 reads an image from the paper when the paper passes through the reading slit 10e2, and generates electronic data (image data) of the image.

  Alternatively, when a sheet is set on the platen glass 10e3, the scanner 10e1 scans the platen glass 10e3 to read an image from the sheet and generate image data.

  A so-called one-line scanner is used as the scanner 10e1. The one-line scanner has a light source and a line sensor for each color of RGB (Red Green Blue), and when one line is read by the line sensor, one color of RGB is turned on and turned on for each line. Switch colors in order. In the following description, it is assumed that the coordinate axis in the main scanning direction is the X axis and the coordinate axis in the sub scanning direction is the Y axis.

  The NIC 10g performs communication using a protocol such as TCP / IP (Transmission Control Protocol / Internet Protocol). The modem 10h communicates with a fax terminal using a protocol such as G3.

  The touch panel display 10i displays a screen for giving a message or an instruction to the user, a screen for the user to input processing instructions and conditions, a screen showing the processing result of the CPU 10a, and the like. Moreover, the position which the user touched with the finger | toe is detected, and the signal which shows a detection result is transmitted to CPU10a.

  The operation key panel 10j is a so-called hardware keyboard, and includes a numeric keypad, a start key, a stop key, a function key, and the like.

  The input image processing unit 10k performs a conversion process for converting the image data of the image read by the scan unit 10e and the image data received by the NIC 10g or the modem 10h into a specific format.

  When printing an image, the output image processing unit 10m performs a conversion process for converting the image data of the image into a specific format.

  The print unit 10f is obtained by the scan unit 10e, the NIC 10g, or the modem 10h, and is appropriately converted by the input image processing unit 10k, and the image data that has been appropriately converted by the output image processing unit 10m is used. Print the image on paper.

  As described above, the read image correction program 10P is a program that reduces the color blur that tends to occur in the one-line scanner as compared with the conventional one. Hereinafter, a mechanism for reducing the color blur as compared with the prior art will be described with reference to FIGS.

  FIG. 4 is a diagram illustrating an example of a functional configuration of the image forming apparatus 1. FIG. 5 is a diagram illustrating an example of the image data 60. FIG. 6 is a diagram illustrating an example of the image 50. FIG. 7 is a diagram illustrating an example of a positional relationship between a read color pixel and a non-read color pixel. FIG. 8 is a diagram illustrating an example of a change in gradation of continuous pixels having the same coordinate in the sub-scanning direction. FIG. 9 is a diagram for explaining an example of a method of calculating the reference gradation. FIG. 10 is a diagram illustrating an example of the reference gradation of the non-read color of each pixel. FIG. 11 is a diagram illustrating an example of a positional relationship between consecutive pixels having the same reading color. FIG. 12 is a diagram illustrating an example of a method of applying the secondary differential filter Fa. FIG. 13 is a diagram illustrating an example of the change feature amount Dh of each pixel. FIG. 14 is a diagram illustrating an example of the correction gradation of the non-read color of each pixel. FIG. 15 is a diagram for explaining a specific example of the calculation of the change feature amount Dh and the correction of the reference gradation. FIG. 16 is a flowchart for explaining an example of the flow of the first gradation determination process. FIG. 17 is a flowchart for explaining an example of the flow of the second gradation determination process. FIG. 18 is a diagram illustrating an example of a result of processing in the black character reading mode by the gradation determination unit 109. FIG. 19 is a diagram illustrating an example of a result of processing in the color image reading mode by the gradation determination unit 109.

  By executing the read image correction program 10P by the CPU 10a, a read color gradation storage unit 101, a read color determination unit 102, a non-read color gradation extraction unit 103, a non-read color reference gradation calculation unit 104, shown in FIG. Functions such as the non-read color reference tone storage unit 105, the read color tone extraction unit 106, the change feature amount calculation unit 107, the non-read color reference tone correction unit 108, the tone determination unit 109, and the image data generation unit 121 Is realized.

  The read color gradation storage unit 101 stores image data 60 as shown in FIG. 5 generated by the scanner 10e1 reading an image from a sheet. Hereinafter, an example in which an image is read from a certain sheet of paper and color bitmap data is stored in the read color gradation storage unit 101 as the image data 60 will be described as an example. The image represented by the image data 60 is referred to as “image 50”. The image 50 is composed of Px × Py pixels.

  As described above, the scanner 10e1 is a one-line scanner. Therefore, as shown in FIG. 6, the image 50 includes a plurality of line images 51 r made up of Px × 1 red pixels, a plurality of line images 51 g made up of Px × 1 green pixels, and Px × 1 pieces. A plurality of line images 51b composed of blue pixels are arranged one by one in the sub-scanning direction in the order of red, green, and blue. Hereinafter, the line images 51r, 51g, and 51b may be collectively referred to as “line images 51”.

  The read color determination unit 102 or the gradation determination unit 109 performs the following processing on the image 50 using the image data 60 stored in the read color gradation storage unit 101.

  The read color discriminating unit 102 pays attention to one line image 51, and the line image 51 is read in any color of red, green, or blue, that is, read by any color light source of the scanner 10e1. Determine whether it was.

  Hereinafter, the noticed line image 51 is referred to as a “note line image 52”. Further, the color of the light source used in reading the target line image 52 is described as “read color”, and the color that is not the read color is described as “non-read color”.

  The non-read color gradation extraction unit 103 extracts the respective gray levels before and after the target line image 52 for each of the two non-read colors.

  The non-read color gradation extraction unit 103 pays attention to the first pixel of the attention line image 52. Hereinafter, the pixel of interest is referred to as a “pixel of interest 52c”. The first non-read color gradation (pixel value) and the second non-read color gradation of the target pixel 52c are extracted immediately before and immediately after the sub-scanning direction, respectively.

  For example, if the target line image 52 is a certain line image 51r and the coordinates of the target pixel 52c are (Xc, Yc), as shown in FIG. 7A, green (first unread color) ), The gradation of the pixel (Xc, Yc-2) and the gradation of the pixel (Xc, Yc + 1) are extracted. Further, as shown in FIG. 7B, the gradation of the pixel of (Xc, Yc-1) and the gradation of the pixel of (Xc, Yc + 2) are set as the gradation of blue (second non-read color). Extract.

The non-reading color reference gradation calculation unit 104 calculates the reference gradations of each of the two non-reading colors in the target pixel 52c using the following expressions (1) and (2).
D1 = Dp × (Yq / 3) + Dq × (Yp / 3) (1)
D2 = Ds × (Yt / 3) + Dt × (Ys / 3) (2)
However, D1 is the reference gradation of the first non-read color. Dp and Dq are the gradation of the first pixel and the gradation of the second pixel extracted by the non-reading color gradation extraction unit 103 of the first non-reading color, respectively. Yp and Yq are the distance between the target pixel 52c and the first pixel and the distance between the target pixel 52c and the second pixel, respectively.

  Similarly, D2 is the reference gradation of the second non-read color. Ds and Dt are, respectively, the gradation of the first pixel and the gradation of the second pixel extracted by the non-reading color gradation extraction unit 103 of the second non-reading color. Ys and Yt are the distance between the target pixel 52c and the first pixel and the distance between the target pixel 52c and the second pixel, respectively.

  For example, when the color (read color) of the pixel of interest 52c is red and the gradation of the pixel of interest 52c and the gradation extracted by the non-read color gradation extraction unit 103 are as shown in FIG. The reference gradations D1 and D2 for blue are represented as shown in FIGS. 9A and 9B. Then, these reference gradations D1 and D2 are calculated by the equations (1) and (2).

  The non-read color gradation extraction unit 103 sequentially pays attention to the remaining pixels (second and subsequent pixels) of the target line image 52, and each of the pixels (target pixel 52c) immediately before and immediately after the sub-scanning direction. The gradation of the first non-reading color (pixel value) and the gradation of the second non-reading color are extracted. Then, the non-reading color reference gradation calculation unit 104 calculates the reference gradations of each of the two non-reading colors in the target pixel 52c according to the expressions (1) and (2).

  Further, the read color determining unit 102 sequentially pays attention to the remaining line images 51 and determines the read color of the line image 51 (target line image 52). Then, the non-read color gradation extraction unit 103 and the non-read color reference gradation calculation unit 104 execute the above-described processing.

  The non-reading color reference tone storage unit 105 calculates the reference tone for each non-reading color of each pixel of each line image 51 calculated by the non-reading color reference tone calculation unit 104 as shown in FIG. It is stored in association with the coordinates of the pixel.

  The read color gradation extraction unit 106 reads the target line image 52 by the read color determination unit 102 before, after, or in parallel with the processing by the non-read color gradation extraction unit 103 and the non-read color reference gradation calculation unit 104. Each time a color is determined, the gradation of three pixels is extracted as follows.

  The read color gradation extraction unit 106 pays attention to the first pixel of the attention line image 52. Hereinafter, the focused pixel is referred to as a “focused pixel 52e”. The gradation (pixel value) of the target pixel 52e is extracted as the gradation D5. Further, the gradations immediately before and immediately after this read color in the sub-scanning direction of the target pixel 52e are extracted as gradations D4 and D6 from the read color gradation storage unit 101 (see FIG. 4). That is, the gradation of the pixel that has returned three in the Y-axis direction (hereinafter referred to as “front pixel 52u”) and the gradation of the pixel that has advanced three (hereinafter referred to as “rear pixel 52v”). Extract.

  For example, when the target line image 52 is the line image 51r and the coordinates of the target pixel 52e are (Xe, Ye), the read color gradation extracting unit 106, as shown in FIG. 11, (Xe, Ye− The gradation of the pixel 3) is extracted as the gradation of the previous pixel 52u, that is, the gradation D4. The gradation of the pixel (Xe, Ye) is extracted as the gradation of the pixel of interest 52e, that is, the gradation D5. The gradation of the pixel (Xe, Ye + 3) is extracted as the rear pixel 52v, that is, the gradation D6.

  It can be said that the gradations D4, D5, and D6 represent a change in gradation in a section before and after the target pixel 52e, that is, in a section from the previous pixel 52u to the rear pixel 52v.

  The reading color gradation extraction unit 106 similarly pays attention to the second and subsequent pixels, and calculates gradations D4, D5, and D6. Further, every time the other line image 51 is noticed, each pixel of the other line image 51 (the noticed line image 52) is noticed, and gradations D4, D5, and D6 are extracted.

  In this way, the read color gradation extraction unit 106 extracts a plurality of sets of gradations D4, D5, and D6.

  The change feature amount calculation unit 107 calculates the change feature amount Dh representing the change feature of the gradation in the section before and after the target pixel 52e, for example, as follows.

As shown in FIG. 12, the change feature amount calculation unit 107 applies a second-order differential filter (Laplacian filter) Fa to the gradations D4, D5, and D6. Thereby, gradations D4 ′, D5 ′, and D6 ′ are obtained. Then, the sum of the gradations D4 ′, D5 ′, and D6 ′ is calculated as the change feature amount Dh. In this example,
Dh = D4 ′ + D5 ′ + D6 ′ = − D4 + 2 · D5-D6
Is calculated.

  As described above, a plurality of sets of gradations D4, D5, and D6 are calculated by the read color gradation extraction unit 106. The change feature amount calculation unit 107 calculates a change feature amount Dh for each group. Thereby, the change feature amount Dh of each group (pixel) as shown in FIG. 13 is calculated.

  The non-reading color reference gradation correcting unit 108 changes the reference gradation of each pixel, that is, the gradation of each of the two non-reading colors (see FIG. 10) stored in the non-reading color reference gradation storage unit 105. Correction is performed by adding the change feature amount Dh of the pixel calculated by the feature amount calculation unit 107.

In other words, if the read color of the pixel is red,
Dgxy ′ = Dgxy + Dhxy
Dbxy ′ = Dbxy + Dhxy
Is calculated. Dgxy ′ is a corrected reference gradation Dg (green reference gradation) of a pixel whose X coordinate is “x” and Y coordinate is “y”. Similarly, Dbxy ′ is the corrected reference gradation Db (blue reference gradation) of the pixel. Dhxy is the change feature amount Dh of the pixel.

Or, if the pixel reading color is green,
Drxy '= Drxy + Dhxy
Dbxy ′ = Dbxy + Dhxy
Is calculated. Drxy ′ is a corrected reference gradation Dr (red reference gradation) of a pixel whose X coordinate is “x” and Y coordinate is “y”.

Or, if the pixel reading color is blue,
Drxy '= Drxy + Dhxy
Dgxy ′ = Dgxy + Dhxy
Is calculated.

  Thereby, the reference gradation after the correction of the non-read color of each pixel as shown in FIG. 14 is calculated. Hereinafter, the corrected reference gradation is referred to as “correction gradation”.

  Here, with reference to FIG. 15, a specific example of processing of the change feature amount calculation unit 107 and the non-read color reference tone correction unit 108 will be described.

  In this specific example, the read color of the target pixel 52e is red, and the gradations D4, D5, and D6 are “130”, “95”, and “120”, respectively, as shown in FIG. . The reference gradations of the two non-read colors of the target line image 52, that is, green and blue, are “210” and “170” as shown in FIG.

  The change feature quantity calculation unit 107 calculates “−60” as the change feature quantity Dh as shown in FIG. 15C by multiplying | 130 95 120 | by the secondary differential filter Fa.

  As shown in FIG. 15D, the non-reading color reference gradation correcting unit 108 adds the change feature amount Dh “−60” to the reference gradations “210” and “170” of green and blue, respectively. Thereby, “150” and “110” are calculated as the correction gradations of green and blue, respectively.

  By the way, in general, according to a one-line scanner, red, green, and blue line images are obtained one by one from three consecutive lines. These three color line images are handled as a set of images.

  However, according to the above-described processing of the read color determination unit 102 or the non-read color reference tone correction unit 108, the read color tone and the two non-read color tones (correction tone) from one line. ) Is obtained. That is, gradations of all three colors are obtained from all three lines. Therefore, as shown in FIGS. 5 and 14, image data having a resolution three times that of the conventional one-line scanner can be obtained.

  Therefore, the gradation determination unit 109 executes processing for matching with the resolution of the conventional one-line scanner in the procedure shown in FIG. Hereinafter, the case where the three line images 51 (51r, 51g, 51b) that are continuous in the order of red, green, and blue are treated as one set will be described as an example, and the gradation determination unit 109 sets each group of red, green, and blue. A method for determining each gradation will be described.

  When the reading mode is the black character reading mode, the gradation determining unit 109 determines the gradation according to the procedure shown in FIG. The “black character reading mode” is a mode for clearly reproducing black characters.

  In FIG. 16, the gradation determination unit 109 pays attention to the first set of line images 51r, 51g, and 51b from which corrected gradations of two non-reading colors are obtained (# 701, # 702).

  Further, the gradation determining unit 109 pays attention to the first coordinate in the main scanning direction (that is, the X coordinate) of each of the noticed line images 51r, 51g, and 51b (# 703, # 704). Then, the saturation of the pixel of the noted X coordinate in each of the noticed line images 51r, 51g, and 51b is calculated (# 705).

  Note that the saturation is represented by a difference between the maximum gradation and the minimum gradation among the red, green, and blue gradations. The gradation of each pixel is stored in the read color gradation storage unit 101 (see FIG. 5) or the non-read color reference gradation storage unit 105 (see FIG. 10).

  Then, the pixel having the lowest saturation among these three pixels is selected (# 706), and the red, green, and blue gradations of the noted X coordinate of the noted group are selected as red, The gradation is determined for each of green and blue (# 707).

  For example, the sub-scanning direction coordinates (that is, the Y coordinate) of the noticed line images 51r, 51g, and 51b are “4”, “5”, and “6”, respectively, and the noticed main scanning direction coordinates are “1”. ”, As shown in FIG. 5, the red, green, and blue gradations of the pixel of the line image 51r, that is, the pixel of the coordinates (1, 4) are“ Dr14 ”,“ Dg14 ′ ”,“ Db14 ′ ”. Further, the red, green, and blue gradations of the pixel of the line image 51g, that is, the pixel of coordinates (1, 5) are “Dr15 ′”, “Dg15”, and “Db15 ′”, respectively, as shown in FIG. is there. The red, green, and blue gradations of the pixel of the line image 51b, that is, the pixel at the coordinates (1, 6) are “Dr16 ′”, “Dg16 ′”, and “Db16”, respectively, as shown in FIG. is there.

  Therefore, in this case, the gradation determination unit 109 calculates the saturation C1 by calculating the difference between the maximum value and the minimum value of the gradations Dr14, Dg14 ′, and Db14 ′, and the gradation Dr15 ′, The saturation C2 is calculated by calculating the difference between the maximum value and the minimum value of Dg15 and Db15 ′, and the difference between the maximum value and the minimum value of the gradations Dr16 ′, Dg16 ′, and Db16 is calculated. Thus, the saturation C3 is calculated.

  Then, the gradation determination unit 109 selects the pixel with the lowest saturation among the saturations C1, C2, and C3, and the red, green, and blue coordinates having the X coordinate of “1” for the set of interest. Are determined as the respective red, green, and blue gradations of the selected pixel (# 707). For example, when the saturation C1 is the lowest, “Dr14”, “Dg14 ′”, and “Db14 ′” are determined.

  The gradation determination unit 109 sequentially pays attention to the second and subsequent coordinates (X = 2, 3,..., Px) in the main scanning direction (No in # 708, # 709), and steps # 704 to # 707. Execute the process.

  Further, the gradation determination unit 109 sequentially pays attention to the second and subsequent sets of line images 51r, 51g, and 51b from which the correction gradations of the two non-read colors are obtained (No in # 710, # 711). , Steps # 703 to # 709 are executed.

  Note that according to the above-described calculation method by the non-reading color reference tone calculation unit 104, the reference tone of the non-reading color is not calculated for the line images 51r, 51g, and 51b at both ends in the sub-scanning direction. Therefore, the correction gradation is not calculated. Therefore, the gradation determination unit 109 may determine the gradation of the read color shown in FIG. 5 without performing the processes of steps # 703 to # 709 for these sets (# 712).

  On the other hand, when the reading mode is the color image reading mode, the gradation determination unit 109 determines the gradation according to the procedure shown in FIG. The “color image reading mode” is a mode for reproducing a color image while suppressing blackening.

  In FIG. 17, the gradation determination unit 109 pays attention to the first set of line images 51r, 51g, and 51b from which corrected gradations of two non-reading colors are obtained (# 721, # 722).

  Further, the gradation determination unit 109 focuses on the first coordinate in the main scanning direction of each of the focused line images 51r, 51g, and 51b (# 723, # 724). Then, for each of red, green, and blue, the average value of the gradations of the noticed line images 51r, 51g, and 51b is calculated (# 725).

  In the example described above, the coordinates of the noticed line images 51r, 51g, and 51b in the sub scanning direction are “4”, “5”, and “6”, respectively, and the noticed coordinates in the main scanning direction are “1”. In this case, the gradation determination unit 109 calculates (Dr14 + Dr15 ′ + Dr16 ′) / 3 as the average value of the red gradation. Similarly, (Dg14 ′ + Dg15 + Dg16 ′) / 3 is calculated as the average value of the green gradation, and (Db14 ′ + Db15 ′ + Db16) / 3 is calculated as the average value of the blue gradation.

  Then, the gradation determination unit 109 calculates the average values of the red, green, and blue gradations calculated in step # 725 for the red, green, and blue gradations that have coordinates of which the X coordinate of the set of interest is “1”. (# 726).

  The gradation determining unit 109 sequentially pays attention to the second and subsequent coordinates (X = 2, 3,..., Px) in the main scanning direction (No in # 727, # 728), and steps # 724 to # 726 are performed. Execute the process.

  Further, the gradation determination unit 109 sequentially pays attention to the second and subsequent sets of line images 51r, 51g, and 51b from which correction gradations of two non-read colors are obtained (No in # 729, # 730). , Steps # 723 to # 728 are executed.

  Then, the gradation determination unit 109 does not perform the processes of Steps # 723 to # 728 for the line images 51r, 51g, and 51b at both ends in the sub-scanning direction, and the gradation of the read color shown in FIG. (# 731).

  As described above, in the black character reading mode, the red, green, and blue gradations are determined so that the saturation is low, and in the color image reading mode, the red, green, and blue gradations are individually averaged. Determine by calculating. Therefore, in the black character reading mode, as shown in FIG. 18, the gradations of the respective colors tend to be closer than in the color image mode. On the other hand, in the color image reading mode, as shown in FIG. 19, the gradations of the respective colors are more easily dispersed than in the black character mode.

  The reading mode may be set in advance by the user. Alternatively, the gradation determination unit 109 may determine and set the characteristics of the image 50 by a known method.

  The image data generation unit 121 generates image data 61 indicating the gray levels of red, green, and blue for each pixel of each group determined by the gray level determination unit 109. Then, the image data 61 is stored or output to another device. Alternatively, the image data 61 is converted into CMYK (cyan magenta yellow black) data, and the image is printed on a sheet by the print unit 10f.

  FIG. 20 is a flowchart for explaining an example of the overall processing flow of processing by the read image correction program 10P.

  Next, an example of the overall processing flow when the image forming apparatus 1 corrects the image 50 will be described with reference to FIG.

  The image forming apparatus 1 executes processing according to the procedure shown in FIG. 20 based on the read image correction program 10P.

  When the image forming apparatus 1 reads an image of a document by the scanner 10 e 1 to generate and store the image data 60, the image forming apparatus 1 executes the following processing based on the image data 60.

  The image forming apparatus 1 determines the reading color for each line image 51 (# 11 in FIG. 20). For each line image 51, the gradations of the two non-reading colors of each pixel are extracted (# 12), and the reference gradations of the two non-reading colors of each pixel are calculated (# 13).

  Before or after or in parallel with the processing of steps # 12 to # 13, the image forming apparatus 1 extracts the gradation of the read color for each line image 51, and calculates the read color change feature amount Dh of each pixel. (# 14, # 15).

  Further, for each pixel, the image forming apparatus 1 corrects the reference gradation of the two non-read colors of the pixel by adding the change feature amount Dh of the pixel (# 16). Thereby, the correction gradation is calculated.

  If the black character reading mode is set (Yes in # 17), the image forming apparatus 1 performs a process for matching the resolution of the one-line scanner by the method shown in FIG. 16 (# 18). On the other hand, when the color image reading mode is set (No in # 17), the image forming apparatus 1 performs processing for matching the resolution of the one-line scanner by the method shown in FIG. 17 (# 19). .

  Then, the image forming apparatus 1 generates the image data 61 using the result of the process of step # 18 or # 19 (# 20).

  According to this embodiment, an image read by a one-line scanner can be corrected more suitably than in the past. In particular, according to the black character mode, color blur can be reduced as compared with the conventional case, and black characters can be clearly reproduced. Further, according to the color image mode, the color price can be reduced and the color characters can be clearly reproduced.

  FIG. 21 is a flowchart for explaining an example of the flow of the reference gradation correction process. FIG. 22 is a diagram illustrating an example of the positional relationship between the target pixel 52w and the pixels before and after the target pixel 52w. FIG. 23 is a diagram illustrating an example of a method of applying the secondary differential filter Fb.

  In this embodiment, the non-reading color reference gradation correcting unit 108 corrects the reference gradation by adding the change feature amount Dh unconditionally. However, this method may overcorrect. Therefore, for example, the reference gradation may be corrected by the procedure shown in FIG. Hereinafter, a modification example of correction will be described by taking a case where the reference gradation of a certain pixel of interest 52w is a target as an example.

  The non-read color reference gradation correction unit 108 extracts the gradation of the pixel of interest 52w from the read color gradation storage unit 101 (# 131). Further, gradations of some pixels around the pixel of interest 52w in the sub-scanning direction are extracted (# 132). For example, as shown in FIG. 22, the gradation of three pixels each before and after in the sub-scanning direction is extracted. Then, among the gradations extracted in steps # 131 and # 132 (seven gradations in the example of FIG. 22), the maximum gradation (hereinafter referred to as “MAX value”) and the minimum floor. Key (hereinafter referred to as “MIN value”) is selected.

  In parallel with or before or after the processing of Steps # 131 to # 133, the non-reading color reference tone correction unit 108 changes the feature amount of the target pixel 52w to the reference tone of each of the two non-read colors of the target pixel 52w. By adding Dh, a first addition value and a second addition value are calculated (# 134). For example, when the read color of the target pixel 52w is red, the sum of the green reference gradation and the change feature amount Dh is calculated as the first addition value, and the blue reference gradation and the change feature amount Dh are calculated. The sum is calculated as the second added value.

  Then, a value obtained by correcting the reference gradation of each of the two non-read colors, that is, a corrected gradation is determined as follows using the MAX value and the MIN value.

  If the first addition value is larger than the MAX value (Yes in # 135), the non-read color reference tone correction unit 108 determines the first non-read color correction tone as the MAX value (# 136). ).

  Alternatively, when the first addition value is less than the MIN value (No in # 135, Yes in # 137), the non-read color reference tone correction unit 108 sets the correction tone of the first non-read color. The MIN value is determined (# 138).

  Alternatively, when the first addition value is equal to or smaller than the MAX value and equal to or greater than the MIN value (No in # 135, No in # 137), the non-reading color reference gradation correcting unit 108 performs the first non-reading. The color correction gradation is determined as the first addition value (# 139).

  For the second non-read color correction gradation, similarly to the first non-read color correction gradation, the non-read color reference gradation correction unit 108 sets the second added value to the MAX value or MIN. It determines by comparing with a value (# 135- # 139).

  Thus, by setting the MAX value and the MIN value to the upper limit and the lower limit of correction, it is possible to suppress the occurrence of color blur on the contrary by the correction.

  In the present embodiment, each unit shown in FIG. 4 is realized by executing the read image correction program 10P by the CPU 10a. You may implement | achieve by the circuit of.

  In the present embodiment, after the image data 60 is obtained, that is, after the entire data of the image 50 is obtained, the read color determination unit 102 or the non-read color reference gradation correction unit 108 performs the correction process as described above. Went. However, the correction process may be performed in parallel with the scan process by the scanner 10e1.

  In the present embodiment, as shown in FIG. 12, the change feature amount calculation unit 107 uses the change feature amount Dh of the target pixel 52e as the read color gradation (gradation D5) of the target pixel 52e and the subordinate of the target pixel 52e. The calculation was performed based on the gradations (gradation levels D4 and D6) of the pixels immediately before and after the read color in the scanning direction. However, it may be calculated using gradations of peripheral pixels other than the immediately preceding and immediately following pixels.

  For example, as illustrated in FIG. 23, the change feature amount calculation unit 107 extracts the gradation (gradation D52) of the target pixel 52e. Further, the gradations (gradation levels D41, D42, D43, D51, D53, D61, D62, D63) of the pixels closest to the target pixel 52e that have the same reading color as the target pixel 52e in each of the eight directions are extracted. To do. A secondary differential filter Fb is applied to the extracted gradation. Thereby, gradations D41 ', D42', D43 ', D51', D52 ', D53', D61 ', D62', and D63 'are obtained.

  Then, the change feature amount calculating unit 107 calculates the sum of these gradations as the change feature amount Dh.

In the example of FIG.
Dh = D41 ′ + D42 ′ + D43 ′ + D51 ′ + D52 ′ + D53 ′ + D61 ′
+ D62 '+ D63'
= -D41-D42-D43-D51 + 8.D52-D53-D61
-D62-D63
Is calculated.

  Similarly, the non-reading color reference gradation calculating unit 104 determines the correction gradation of the non-reading color not only for the gradation of the two pixels before and after the pixel of interest 52c (see FIG. 11) but also for the eight pixels around the eight directions. It may be calculated on the basis of the gradation. In this case, the correction gradation may be calculated by a known method of linear interpolation.

  In addition, the configuration of the entire image forming apparatus 1 or each unit, the content of processing, the order of processing, the configuration of data, and the like can be appropriately changed in accordance with the spirit of the present invention.

1 Image forming device (read image correction device)
103 Non-read color gradation extraction unit (non-read color gradation calculation means)
107 Change feature amount calculation unit (feature amount calculation means)
108 Non-reading color reference gradation correction unit (correction means)
109 Gradation determination unit (resolution adjustment means)
10e1 scanner (1 line scanner)
50 images (read images)
Dh change feature (feature)

Claims (9)

A read image correction apparatus that corrects a read image that is an image read by a one-line scanner while sequentially turning on a light source of each of a plurality of colors for each line,
A feature amount representing a feature of change of the target pixel as a correction target in the read image from the periphery of the target pixel is a gradation of the target pixel obtained by the one-line scanner. Read by the one-line scanner when the reading light source used for reading the target pixel among the light sources of the plurality of colors in the read image is turned on, and A feature amount calculating means for calculating by a second reading gradation that is a gradation of a pixel adjacent to the target pixel in the sub-scanning direction;
A non-reading color gradation that is a gradation of a non-reading color that is a color different from a color corresponding to the reading light source among the plurality of colors of the target pixel. When the light source of the non-read color among the light sources of each color is lit, the third read gradation of each of the plurality of pixels read by the one-line scanner and close to the target pixel in the sub-scanning direction Non-reading color gradation calculating means for calculating based on
Correction means for calculating a correction gradation by correcting the non-read color gradation based on the feature amount;
A read image correction apparatus comprising: The feature amount calculating means calculates the feature amount by adding a value obtained by applying a second-order differential filter to the first reading gradation and the second reading gradation,
The correction means calculates the correction gradation by adding the feature amount to the non-read color gradation;
The read image correction apparatus according to claim 1. The feature amount calculating means calculates the feature amount by adding a value obtained by applying a second-order differential filter to the first reading gradation and the second reading gradation,
When the feature value is a positive value, the correction unit determines the absolute value of the difference between the maximum value of the second reading gradation and the third reading gradation and the non-reading color gradation. The correction gradation is calculated by adding the smaller one of the feature amounts to the non-read color gradation, and if the correction amount is a negative value, the second reading gradation and the second reading gradation are calculated. By subtracting the smaller of the absolute value of the difference between the minimum value of the three reading gradations and the non-reading color gradation and the absolute value of the feature amount from the non-reading color gradation Calculating the correction gradation;
The read image correction apparatus according to claim 1. The plurality of colors are three primary colors, a first primary color, a second primary color, and a third primary color,
The non-reading color gradation calculating unit calculates the non-reading color gradation for each of two of the three primary colors;
The correction means calculates the correction gradation for each of two of the three primary colors;
The read image correction apparatus according to claim 1. As a common gradation of a set of three pixels continuously arranged in the sub-scanning direction in the read image, the first gradation and the non-read color gradation of each of the three pixels. A first common gradation that is the gradation of the first primary color is calculated based on the first reading gradation and the non-reading color gradation of the first primary color, and the first reading of the second primary color is performed. A second common gradation that is the gradation of the second primary color is calculated based on the gradation and the non-reading color gradation, and the first reading gradation and the non-reading color gradation of the third primary color are calculated. A resolution adjusting means for calculating a third common gradation which is a gradation of the third primary color based on
The read image correction apparatus according to claim 4. The resolution adjusting unit calculates the first common gradation by selecting the lowest saturation among the first reading gradation and the non-reading color gradation of the first primary color, The second common gradation is calculated by selecting the first reading gradation and the non-reading color gradation of the second primary color having the lowest saturation, and the third common gradation is calculated. A tone is calculated by selecting the lowest saturation of the first read gradation and the non-read color gradation of the third primary color;
The read image correction apparatus according to claim 5. The resolution adjusting unit calculates an average value of the first reading gradation and the non-reading color gradation of the first primary color as the first common gradation, and the second common gradation as the second common gradation. An average value of the first reading gradation and the non-reading color gradation of the primary color is calculated, and the average of the first reading gradation and the non-reading color gradation of the third primary color is calculated as the third common gradation. Calculate the value,
The read image correction apparatus according to claim 5. A read image correction method for correcting a read image that is an image read by a one-line scanner while sequentially turning on light sources of a plurality of colors for each line.
A feature amount representing a feature of change of the target pixel as a correction target in the read image from the periphery of the target pixel is a gradation of the target pixel obtained by the one-line scanner. Read by the one-line scanner when the reading light source used for reading the target pixel among the light sources of the plurality of colors in the read image is turned on, and And a second reading gradation that is a gradation of a pixel adjacent to the target pixel in the sub-scanning direction,
A non-reading color gradation that is a gradation of a non-reading color that is a color different from a color corresponding to the reading light source among the plurality of colors of the target pixel. When the light source of the non-read color among the light sources of each color is lit, the third read gradation of each of the plurality of pixels read by the one-line scanner and close to the target pixel in the sub-scanning direction Based on
Calculating a correction gradation by correcting the non-read color gradation based on the feature amount;
A read image correction method characterized by the above. A computer program used for a computer that corrects a read image, which is an image read by a one-line scanner, while sequentially turning on a light source of each of a plurality of colors for each line,
In the computer,
A feature amount representing a feature of change of the target pixel as a correction target in the read image from the periphery of the target pixel is a gradation of the target pixel obtained by the one-line scanner. Read by the one-line scanner when the reading light source used for reading the target pixel among the light sources of the plurality of colors in the read image is turned on, and A second reading gradation that is a gradation of a pixel adjacent to the target pixel in the sub-scanning direction is executed,
A non-reading color gradation that is a gradation of a non-reading color that is a color different from a color corresponding to the reading light source among the plurality of colors of the target pixel. When the light source of the non-read color among the light sources of each color is lit, the third read gradation of each of the plurality of pixels read by the one-line scanner and close to the target pixel in the sub-scanning direction Based on the calculation process,
Executing a process of calculating a correction gradation by correcting the non-read color gradation based on the feature amount;
A computer program characterized by the above.

Images