JP2008005192A - Image processor, image processing method, and image processing program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processor, an image processing method, and an image processing program that can precisely make a chromatic/achromatic color decision on image data obtained by reading a document without being affected by a color shift etc. <P>SOLUTION: The image processor 10 comprises: a read unit 20 which outputs R, G, and B values by pixels; a test pattern 18 having black lines 40; a color shift correction coefficient calculator 60 which calculates color shift correction coefficients by the R, G, and B components of image data obtained by reading the test pattern 18; the maximum detector 80 which detects the maximum among the respective R, G, and B values of a document image; and a mean value calculator 84 which weights the respective R, G, and B values of the document image by respective R, G, and B color shift correction coefficients to calculate a weighted mean value thereof. Then the difference between the RGB maximum and the RGB weighted mean value is calculated and compared with a predetermined threshold to decide whether an object pixel has a chromatic or an achromatic color. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image processing device, an image processing method, and an image processing program, and more particularly, to an image processing device, an image processing method, and an image processing program that determine whether read image data is a chromatic color or an achromatic color.

  In order to improve the reproducibility of image data, a color image processing apparatus that determines whether read image data is chromatic or achromatic is known. In Patent Document 1, the largest value and the smallest value among the R value, the G value, and the B value are obtained for each pixel of the read image data, and a difference value between the maximum value and the minimum value is calculated. It is disclosed to determine whether image data is a chromatic color or an achromatic color by comparing the value with a predetermined threshold value.

JP 2004-88663 A (paragraphs [0043] [0044] and FIG. 5)

  If the image data is a photographic image, there is no problem with the above method. However, when the image data is a character image, the above method causes a problem. This is because, in a color image processing apparatus represented by a color scanner, color misregistration may occur at the edge of a character due to processing characteristics. Image data in which color misregistration has occurred has a larger density difference between RGB than image data in which color misregistration has not occurred. Therefore, since the conventional method simply takes the difference value between the maximum value and the minimum value, it is detrimentally affected by the variation in density difference due to this color shift. There is a concern that a chromatic pixel may be misjudged as a chromatic color.

  The present invention has been made to solve the above-described problems, and can execute chromatic / achromatic color determination of image data obtained by reading a document with high accuracy without being affected by color misregistration or the like. An object is to provide a possible image processing apparatus, an image processing method, and an image processing program.

  In order to solve the above problems, the image processing apparatus of the present invention includes means for determining whether a pixel is a chromatic color or an achromatic color based on a plurality of color component values included in the pixel in the image and a predetermined threshold value. An image processing apparatus, a test pattern for acquiring data used for image correction, a correction coefficient calculation unit that calculates a correction coefficient for each RGB component of a pixel based on the data, and each RGB of a pixel in an image A weighted average value calculating unit that weights each value with the correction coefficient and calculates an average value thereof, a difference value between a maximum value among the RGB values of the pixel and the average value, and a predetermined threshold value And a color determination unit that determines whether the pixel is a chromatic color or an achromatic color.

  In this case, it is possible to further include a character image determination unit that determines whether or not the pixels in the image are character images. And when the said character image discrimination | determination part discriminate | determines the said pixel as a character image, the said color determination part can make the difference value used for color determination into the difference value of the said maximum value and the said average value. On the other hand, when the character image determination unit determines that the pixel is an image other than a character image, the color determination unit determines a difference value used for color determination between the maximum value and the minimum value among the RGB values. It can also be a difference value.

  The test pattern may include a black line having a predetermined width. In this case, the correction coefficient calculation unit can also obtain each RGB average value in a predetermined small area including the edge of the black line, and calculate each RGB correction coefficient from each average value.

  Further, the black line can include a black line provided substantially in parallel with the sub-scanning direction, or the black line can include a black line provided substantially in parallel with the main scanning direction. . Furthermore, a plurality of the black lines can be provided in the main scanning direction. Furthermore, the width of the black line may be set to a width corresponding to 5 pixels, and the small area may be an 8 × 8 pixel area.

  The test pattern may be placed between the white reference plate position and the document reading start position. In addition, the color determination unit can set the predetermined threshold corresponding to the average value or an arithmetic average value of the RGB values.

  The image processing method of the present invention is an image processing method for determining whether a pixel is a chromatic color or an achromatic color based on a plurality of color component values included in the pixel in the image and a predetermined threshold, and a test pattern Is calculated for each RGB component of the pixel, and each RGB value of the pixel in the image is weighted with the correction coefficient to calculate an average value thereof, and each RGB value of the pixel in the image is calculated. A difference value between the maximum value of the values and the average value is calculated, and the difference value is compared with a predetermined threshold value to determine whether the pixel in the image is a chromatic color or an achromatic color.

  In this case, it is possible to further include a process of determining whether or not the pixels in the image are character images. When the image is determined to be a character image, the difference value may be a difference value between the maximum value and the average value. On the other hand, when the image is determined to be an image other than a character image, the difference value may be a difference value between the maximum value and the minimum value among the RGB values.

  In the correction coefficient calculation process, RGB average values in a predetermined small area including the edge of the black line constituting the test pattern may be obtained, and RGB color misregistration correction coefficients may be calculated from the average values.

  The width of the black line may be set to a width corresponding to 5 pixels, and the small area may be an 8 × 8 pixel area. The predetermined threshold value may be set corresponding to the average value or an arithmetic average value of the RGB values.

  The image processing program of the present invention is an image processing program for determining whether a pixel is a chromatic color or an achromatic color based on a plurality of color component values included in the pixel in the image and a predetermined threshold, and a test pattern A function for calculating a correction coefficient for each RGB component of the pixel, a function for weighting each RGB value of the pixel in the image with the correction coefficient and calculating an average value thereof, A function for calculating a difference value between the maximum value among the RGB values of the pixel and the average value, and comparing the difference value with a predetermined threshold value to determine whether the pixel in the image is chromatic or achromatic. Let the computer execute the function of determining.

  In the present invention, the chromatic / achromatic color determination of a pixel in an image is weighted by a correction coefficient calculated based on a maximum value among RGB values of the pixel and data acquired from a test pattern. This is performed based on a difference value from the average value of each RGB value of the pixel. Therefore, even when color unevenness occurs in the edge portion of a character, the density difference between RGB is large compared to the conventional configuration in which chromatic / achromatic color determination is performed based on the difference value between the maximum value and the minimum value. Never become. Therefore, in the case of the present invention, for example, an achromatic pixel is not erroneously determined as a chromatic color in the character peripheral portion. That is, according to the present invention, the chromatic / achromatic color determination of the target pixel can be performed with high accuracy.

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

  FIG. 1 is an example of a side view of an image processing apparatus 10 according to an embodiment of the present invention. The image processing apparatus 10 includes a white reference plate 12 for shading correction and a contact glass 16 on which a document 14 is placed on an upper portion of the casing. Between the white reference plate 12 and the contact glass 16, a test pattern 18 to be described later for calculating a color misregistration correction coefficient is provided. Further, the image processing apparatus 10 includes a reading unit 20 that is movable in the sub-scanning direction Y inside the housing. The reading unit 20 includes a light source 22, an image sensor 24 such as a CCD (Charge Coupled Device), an optical system 26 that guides reflected light emitted from the light source 22 and reflected by the document 14, and the like to the image sensor 24, and an image processing unit. 28. Details of the image processing unit 28 will be described later. The reading unit 20 moves along the guide rail 34 in the sub-scanning direction Y by a belt 32 that rotates under the drive of a motor 30. In the present embodiment, the reading unit 20 includes a color image sensor 24 and outputs R (red: Red), G (green: Green), and B (blue: Blue) values for each pixel. And

  FIG. 2 is an example of a schematic diagram of a test pattern 18 for calculating a color misregistration correction coefficient. The test pattern 18 has substantially the same width as the white reference plate 12 and the contact glass 16 in the main scanning direction X. The test pattern 18 includes a black line 40 at a substantially central portion of each of the regions 1 to 5 in the main scanning direction X.

  FIG. 3 is an example of an enlarged view of the black line 40. The black line 40 has a width of about 5 pixels in the main scanning direction X and a length of about 5 mm in the sub-scanning direction Y.

  FIG. 4 is an example of a control block diagram of the image processing unit 28. The image processing unit 28 includes an image input unit 50, a shading correction unit 52, a γ correction unit 54, a character photograph determination unit 56, a color determination unit 58, and a color misregistration correction coefficient calculation unit 60. The image input unit 50 inputs image data from the image sensor 24. The shading correction unit 52 performs shading correction on the image data based on the white reference data obtained by reading the white reference plate 12. The γ correction unit 54 performs γ correction on the image data. The character photograph determination unit 56 determines whether each pixel of the image data is a character image or a photograph image, and outputs the character photograph determination result to the color determination unit 58. The color determination unit 58 determines whether each pixel of the image data is a chromatic color or an achromatic color, and outputs a determination result. The color misregistration correction coefficient calculation unit 60 calculates a color misregistration correction coefficient for each RGB component so as to be less affected by the color misregistration when the color determination unit 58 performs color determination.

  FIG. 5 is an example of a control block diagram of the color misregistration correction coefficient calculation unit 60. The color misregistration correction coefficient calculation unit 60 includes a small area setting unit 70, an average value calculation unit 72, and a color misregistration correction coefficient determination unit 74. The small region setting unit 70 sets a predetermined small region including both end edges in the width direction of the black line 40 from the image data obtained by reading the test pattern 18 (see FIG. 3). In this case, the small region is an 8 × 8 pixel region that is slightly larger than 5 pixels that is the width of the black line 40. The average value calculation unit 72 calculates each average value for each of the R, G, and B components in the 8 × 8 pixel region, and outputs these to the color misregistration correction coefficient determination unit 74. The color misregistration correction coefficient determination unit 74 determines R, G, and B color misregistration correction coefficients Kr, Kg, and Kb from the input average values. In this case, the color misregistration correction coefficient determination unit 74 determines the correction coefficients Kr, Kg, and Kb with reference to any value of RGB. For example, when the average value of R is 27, the average value of G is 30, and the average value of B is 33, and G is used as a reference, the G color misregistration correction coefficient Kg is 1.0. The R color misregistration correction coefficient Kr is 27/30 = 0.9, and the B color misregistration correction coefficient Kb is 33/30 = 1.1.

  As described above, the test pattern 18 includes the black line 40 for each of the regions 1 to 5 in the main scanning direction. Therefore, R, G, and B color misregistration correction coefficients Kr, Kg, and Kb can be calculated for each of the regions 1 to 5. By doing so, even if the reading unit 20 has a variation in reading accuracy in the main scanning direction, a color misregistration correction coefficient corresponding to the variation is obtained, so that a more accurate color determination is possible.

  In addition, by setting the small area to an 8 × 8 area, it is possible to maintain high accuracy of the color misregistration correction coefficients Kr, Kg, and Kb while keeping the circuit scale relatively small. Further, by setting the width of the black line 40 to about 5 pixels, the black line 40 can be accommodated in the 8 × 8 region. That is, both end edges of the black line 40 can be detected at once in the 8 × 8 region.

  FIG. 6 is an example of a control block diagram of the color determination unit 58. The color determination unit 58 includes a maximum value detection unit 80, a minimum value detection unit 82, an average value calculation unit 84, a first difference value calculation unit 86, a first threshold value generation unit 88, and a first determination unit 90. , A second difference value calculation unit 92, a second threshold value generation unit 94, a second determination unit 96, and an output selection unit 98.

  The maximum value detector 80 detects the largest numerical value (hereinafter referred to as the maximum value RGBmax) among the R value, G value, and B value of the pixels obtained by the reading unit 20 reading the document 14. To do. The minimum value detector 82 detects the smallest numerical value (hereinafter referred to as the minimum value RGBmin) of the R value, G value, and B value of the pixels obtained by the reading unit 20 reading the document 14. To do.

  The average value calculation unit 84 inputs RGB values of pixels obtained by the reading unit 20 reading the document 14 and RGB color misregistration correction coefficients Kr, Kg, and Kb output from the color misregistration correction coefficient calculation unit 60. . The average value calculation unit 84 calculates a simple average value (arithmetic average value) RGBave1 of the input RGB values. The arithmetic mean value RGBave1 is calculated by the following mathematical formula (1).

RGBave1 = (R + G + B) / 3 (1)
The calculated arithmetic mean value RGBave1 is output to the first threshold value generator 88.

  The average value calculation unit 84 weights the input RGB values with the color misregistration correction coefficients Kr, Kg, and Kb to calculate a weighted average value (weighted average value) RGBave2. The weighted average value RGBave2 is calculated by the following mathematical formula (2).

RGBave2 = (R · Kr + G · Kg + B · Kb) / 3 (2)
The calculated weighted average value RGBave2 is output to the second difference detection unit 92 and the second threshold value generation unit 94.

  The first difference value calculation unit 86 calculates a difference value DIF1 (= RGBmax−RGBmin) between the maximum value RGBmax and the minimum value RGBmin and outputs the difference value DIF1 (= RGBmax−RGBmin) to the first determination unit 90. The first threshold generation unit 88 generates a first determination threshold THL1 that is a determination criterion in the chromatic / achromatic color determination performed by the first determination unit 90, and outputs the first determination threshold THL1 to the first determination unit 90. The first determination unit 90 compares the difference value DIF1 with the first determination threshold value THL1, determines whether the target pixel is a chromatic color or an achromatic color, and outputs the first determination result to the output selection unit 98. For example, when the comparison result is DIF1 ≧ THL1, the first determination unit 90 determines the target pixel as a chromatic color, and when the comparison result is DIF <THL1, the first determination unit 90 sets the target pixel as an achromatic color. Is determined.

  The second difference value calculation unit 92 calculates a difference value DIF2 (= RGBmax−RGBave2) between the maximum value RGBmax and the weighted average value RGBave2 and outputs the difference value DIF2 (= RGBmax−RGBave2) to the second determination unit 96. The second threshold generation unit 94 generates a second determination threshold THL2 that is a determination criterion in the chromatic / achromatic color determination performed by the second determination unit 96 and outputs the second determination threshold THL2 to the second determination unit 96. The second determination unit 96 compares the difference value DIF2 with the second determination threshold value THL2, determines whether the target pixel is a chromatic color or an achromatic color, and outputs the first determination result to the output selection unit 98. For example, when the comparison result is DIF2 ≧ THL2, the second determination unit 96 determines that the target pixel is a chromatic color, and when the comparison result is DIF2 <THL2, the second determination unit 96 sets the target pixel to an achromatic color. Is determined.

  The output selection unit 98 inputs the first determination result from the first determination unit 90 and inputs the second determination result from the second determination unit 96. Further, a character photo discrimination result indicating whether the target pixel is a character image or a photo image is input from the character photo discrimination unit 56 to the output selection unit 98. The output selection unit 98 outputs either the first determination result or the second determination result as a color determination result according to the determination result.

  The flowchart in FIG. 7 shows an example of an operation example of the image processing apparatus 10 according to the embodiment of the present invention. First, the user places an original 14 to be read on the contact glass 16 and instructs the image processing apparatus 10 to start reading. Receiving the reading start instruction, the image processing apparatus 10 drives the motor 30 to move the reading unit 20 from the predetermined home position toward the contact glass 16.

  The reading unit 20 that has left the home position stops under the white reference plate 12 and performs gain adjustment for each of R, G, and B (step S1). Next, the reading unit 20 drives the light source 22, the image sensor 24, and the image processing unit 28 to read the white reference plate 12, and obtain white reference data for shading correction (step S2). The white reference data is stored in a predetermined memory in the image processing unit 28. The reading unit 20 stopped under the white reference plate 12 starts to move again toward the contact glass 16 under the drive of the motor 30. When the reading unit 20 comes under the test pattern 18, the test pattern 18 is read by the image sensor 24 (step S3). The test pattern image converted into an electrical signal by the image sensor 24 is input to the image processing unit 28. The input test pattern image is converted from an analog signal into a digital signal, and the test pattern image data converted into the digital signal is subjected to shading correction in the shading correction unit 52 and then subjected to γ correction in the γ correction unit 54. The test pattern image data subjected to these processes is input to the correction coefficient calculation unit 60. The small area setting unit 70 sets an 8 × 8 area including both end edges in the width direction of the black vertical lines 40 from the entire test pattern image data (see FIG. 3). The average value calculation unit 72 calculates the average value for each of the R, G, and B components in the 8 × 8 pixel region, and outputs these to the color misregistration correction coefficient determination unit 74. The color misregistration correction coefficient determination unit 74 calculates R, G, and B color misregistration correction coefficients Kr, Kg, and Kb from the input average values (step S4). In this case, the color misregistration correction coefficients Kr, Kg, Kb are calculated for each of the regions 1 to 5, and the calculated R, G, B color misregistration correction coefficients Kr, Kg, Kb are output to the color determination unit 58 as they are. Or stored in a memory (not shown) in the color misregistration correction coefficient calculation unit 60. Eventually, the reading unit 20 that has reached the document reading start position of the contact glass 16 under the drive of the motor 30 activates the light source 22, the image sensor 24, and the image processing unit 28 to read the document 14 (step S5).

  The operation of the image processing unit 28 during document reading will be described. The document image input to the image processing unit 28 is converted from an analog signal into a digital signal. The document image data converted into the digital signal is subjected to shading correction in the shading correction unit 52 and then subjected to γ correction in the γ correction unit 54. Is given. The document image data subjected to the γ correction is input to the character / photograph determination unit 56. Incidentally, the image data after γ correction in FIG. 4 is also input to the color misregistration correction coefficient calculation unit 60. The color misregistration correction coefficient calculation unit 60 basically processes only the image data of the test pattern 18 and does not process the image data of the document 14. Accordingly, the original image data input to the color misregistration correction coefficient calculation unit 60 during document reading is not processed at all, or, for example, the γ correction unit 54 sends the image data to the color misregistration correction coefficient calculation unit 60 during document reading. It is also possible not to output.

  The character photograph determination unit 56 determines whether each pixel of the input document image data is a photograph image or a character image. Then, the character photo discrimination unit 56 outputs the character photo discrimination result and the RGB values of the original image data subjected to each correction process to the color determination unit 58.

  The color determination unit 58 receives R, G, and B color misregistration correction coefficients Kr, Kg, and Kb calculated by the color misregistration correction coefficient calculation unit 60, in addition to the character photo discrimination result and the RGB values of the document image data. Has been. The RGB values of the document image data are respectively input to the maximum value detection unit 80, the minimum value detection unit 82, and the average value calculation unit 84, and the RGB color misregistration correction coefficients Kr, Kg, and Kb are input to the average value calculation unit 84. To do. On the other hand, the character photo discrimination result from the character photo discrimination unit 56 is input to the output selection unit 98.

  The maximum value detecting unit 80 detects the maximum value RGBmax of each input RGB value. The minimum value detector 82 detects the minimum value RGBmin of each RGB value. First, the average value calculation unit 84 calculates the arithmetic average value RGBave1 for each RGB value by the above-described equation (1). Secondly, the average value calculation unit 82 calculates the weighted average value RGBave2 according to Equation (2) described above for each RGB value. The arithmetic average value RGBave1 is output to the first threshold value generator 88, and the weighted average value RGBave2 is output to the second difference value calculator 92 and the second threshold value generator 94.

  The first difference value calculation unit 86 calculates a difference value DIF1 between the maximum value RGBmax and the minimum value RGBmin. The first determination unit 90 determines whether the target pixel is a chromatic color or an achromatic color by comparing the difference value DIF1 with the first determination threshold value THL1 input from the first threshold value generation unit 88, and determines the first determination result. The data is output to the output selection unit 98.

  The second difference value calculation unit 92 calculates a difference value DIF2 between the maximum value RGBmax and the weighted average value RGBave2. The second determination unit 96 determines whether the target pixel is a chromatic color or an achromatic color by comparing the difference value DIF2 with the second determination threshold value THL2 input from the second threshold value generation unit 94, and determines the second determination result. The data is output to the output selection unit 98.

  The output selection unit 98 outputs the first determination result as a color determination result when the character photo determination result is a determination result with the target pixel as a photograph, and the character photo determination result is a determination result with the target pixel as a character. If there is, the second determination result is output as the color determination result.

  As described above, when the target pixel is determined to be a character image, the image processing apparatus 10 and the image processing method thereof according to the present embodiment obtain the chromatic / achromatic color determination of the pixel as the RGB maximum value RGBmax in advance. This is performed based on the difference value DIF2 from the RGB weighted average value RGBave2 weighted by the color misregistration correction coefficients Kr, Kg, and Kb. Therefore, even when color unevenness occurs in the edge portion of the character, the density difference between RGB does not increase compared to the conventional configuration in which color determination is performed based on the difference value between the maximum value and the minimum value. . Therefore, in the case of the present invention, an achromatic pixel is not erroneously determined as a chromatic color in the character peripheral portion. That is, this embodiment can perform the chromatic / achromatic color determination of the target pixel with high accuracy.

  In the above embodiment, the first determination threshold value THL1 may be a fixed threshold value, but can also be a variation threshold value. In this case, as shown in FIG. 8, if the threshold value is changed according to the arithmetic mean value RGBave1, the chromatic / achromatic color determination accuracy when the target pixel is determined to be a photographic image can be improved.

  Further, the second determination threshold value THL2 may be a fixed threshold value, but may be a variation threshold value. In this case, as shown in FIG. 9, if the threshold value changes according to the weighted average value RGBave2, the chromatic / achromatic color determination accuracy when the target pixel is determined to be a character image can be improved.

  As shown in FIG. 3, the color misregistration correction coefficient calculation unit 60 in the above embodiment sets only one small region in the sub-scanning direction of the black line 40, and R, G based on the value of this one small region. , B color misregistration correction coefficients Kr, Kg, Kb are obtained, but the method of obtaining the correction coefficients is not limited to this. For example, as shown in FIG. 10, after setting a plurality of small areas A1 to A3 on the black line 40 and obtaining correction coefficients for each of the small areas A1 to A3, an average value of these correction coefficients is obtained, and this average is obtained. The value can also be a representative correction factor. Since the correction coefficient is determined by sampling a plurality of points of the black line 40 and taking the average value in this way, even if the black vertical line 40 has a blur or noise, the effect is diluted. Can do.

  Further, the test pattern 18 is not limited to the one having the long black line 40 in the sub-scanning direction Y as shown in FIGS. 2 and 3, and for example, the test pattern 18 may have a long black line in the main scanning direction. it can. For example, as shown in FIG. 11, the test pattern may be a test pattern 102 including a black line 100 elongated in the main scanning direction X in addition to the black line 40 elongated in the sub-scanning direction Y. In this case, a small area can be set for each of the black lines 40 and 100, an average value of correction coefficients obtained for each small area can be calculated, and this average value can be used as a representative correction coefficient. By doing so, it is possible to absorb variations in image quality in each scanning direction and calculate more appropriate color misregistration correction coefficients Kr, Kg, and Kb.

  Moreover, in the said embodiment, although the test pattern 18 is provided between the white reference board 12 and the contact glass 16, the arrangement position of the test pattern 18 is not limited to the said place. For example, the test pattern 18 may be disposed at a predetermined position in the direction opposite to the contact glass 16 when viewed from the white reference plate 12. By adopting such an arrangement, for example, an image processing apparatus including an ADF (Auto Document Feeder) in the opposite direction to the contact glass 16 can be handled.

  Further, the width / length of the black line 40 and the size of the small area can be arbitrarily changed.

  Further, the above-described image processing unit 28 includes a CPU (Central Processing Unit) and a memory, stores a program for realizing the above functions in the memory, and the CPU accesses the memory to execute the program. You can also

It is an example of the side view of the image processing apparatus which concerns on embodiment of this invention. FIG. 2 is an example of a schematic diagram of a color misregistration correction coefficient calculation test pattern provided in the image processing apparatus shown in FIG. 1. It is an example of the enlarged view of the black line which comprises the test pattern shown in FIG. FIG. 2 is an example of a control block diagram of an image processing unit constituting the image processing apparatus shown in FIG. 1. FIG. 5 is an example of a control block diagram of a color misregistration correction coefficient calculation unit shown in FIG. 4. It is an example of the control block diagram of the color determination part shown in FIG. 2 is an example of a flowchart illustrating an operation example of the image processing apparatus illustrated in FIG. 1. It is an example of the graph which shows the relationship between the 1st determination threshold value at the time of making a 1st determination threshold value into a fluctuation threshold value, and an arithmetic mean value. It is an example of the graph which shows the relationship between the 2nd determination threshold value when a 2nd determination threshold value is made into a fluctuation | variation threshold value, and a weighted average value. The other setting example of the small area setting in a black vertical line is shown. The other example of a test pattern is shown.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Image processing apparatus 12 White reference board 14 Document 16 Contact glass 18, 102 Test pattern 20 Reading unit 40, 100 Black line 56 Character photograph discrimination part 58 Color judgment part 60 Color shift correction coefficient calculation part 80 Maximum value detection part 82 Minimum value Detection unit 84 Average value calculation unit 86 First difference value calculation unit 90 First determination unit 92 Second difference value calculation unit 96 Second determination unit 98 Output selection unit

Claims (22)

An image processing apparatus comprising means for determining whether a pixel is a chromatic color or an achromatic color based on a plurality of color component values included in the pixel in the image and a predetermined threshold,
A test pattern for acquiring data used for image correction;
A correction coefficient calculation unit that calculates a correction coefficient for each RGB component of the pixel based on the data;
A weighted average value calculating unit that weights each RGB value of the pixel in the image with the correction coefficient and calculates an average value thereof;
A color determination unit that determines whether the pixel is a chromatic color or an achromatic color by comparing a difference value between a maximum value among the RGB values of the pixel and the average value and a predetermined threshold value. An image processing apparatus.   The image processing apparatus according to claim 1, further comprising a character image determination unit that determines whether or not a pixel in the image is a character image.   The said color determination part makes the difference value used for color determination into the difference value of the said maximum value and the said average value, when the said character image determination part discriminate | determines the said pixel as a character image. 2. The image processing apparatus according to 2.   When the character image determination unit determines that the pixel is an image other than a character image, the color determination unit determines a difference value used for color determination as a difference value between the maximum value and the minimum value among the RGB values. The image processing apparatus according to claim 2, wherein the image processing apparatus is an image processing apparatus.   The image processing apparatus according to claim 1, wherein the test pattern includes a black line having a predetermined width.   6. The correction coefficient calculation unit obtains each RGB average value in a predetermined small area including the black line edge, and calculates each RGB correction coefficient from each average value. Image processing device.   The image processing apparatus according to claim 5, wherein the black line includes a black line provided substantially parallel to the sub-scanning direction.   The image processing apparatus according to claim 5, wherein the black line includes a black line provided substantially parallel to the main scanning direction.   The image processing apparatus according to claim 7, wherein a plurality of the black lines are provided in the main scanning direction.   The image processing apparatus according to claim 5, wherein a width of the black line is set to a width corresponding to 5 pixels, and the small region is an 8 × 8 pixel region.   The image processing apparatus according to claim 1, wherein the test pattern is installed between a white reference plate position and a document reading start position.   The image processing apparatus according to claim 1, wherein the color determination unit sets the predetermined threshold corresponding to the average value.   The image processing apparatus according to claim 1, wherein the color determination unit sets the predetermined threshold corresponding to an arithmetic average value of the RGB values. An image processing method for determining whether a pixel is a chromatic color or an achromatic color based on a plurality of color component values included in the pixel in the image and a predetermined threshold,
Read the test pattern, calculate the correction coefficient for each RGB component of the pixel,
Each of the RGB values of the pixels in the image is weighted with the correction coefficient, and an average value thereof is calculated.
Calculating a difference value between the maximum value of the RGB values of the pixels in the image and the average value,
An image processing method comprising: determining whether a pixel in the image is a chromatic color or an achromatic color by comparing the difference value with a predetermined threshold value.   The image processing method according to claim 14, further comprising a process of determining whether or not a pixel in the image is a character image.   The image processing method according to claim 15, wherein when the image is determined to be a character image, the difference value is a difference value between the maximum value and the average value.   The difference value between the maximum value and the minimum value among the RGB values is determined when the image is determined to be an image other than a character image. Image processing method.   In the correction coefficient calculation process, each RGB average value in a predetermined small area including a black line edge constituting the test pattern is obtained, and each RGB color misregistration correction coefficient is calculated from each average value. The image processing method according to claim 14.   19. The image processing method according to claim 18, wherein a width of the black line is set to a width corresponding to 5 pixels, and the small region is an 8 × 8 pixel region.   The image processing method according to claim 14, wherein the predetermined threshold is set corresponding to the average value.   The image processing method according to claim 14, wherein the predetermined threshold is set in correspondence with an arithmetic average value of the RGB values. An image processing program for determining whether a pixel is a chromatic color or an achromatic color based on a plurality of color component values included in the pixel in the image and a predetermined threshold,
A function of reading a test pattern and calculating a correction coefficient for each RGB component of the pixel;
A function of weighting each RGB value of the pixel in the image with the correction coefficient and calculating an average value thereof;
A function of calculating a difference value between the maximum value of the RGB values of the pixels in the image and the average value;
An image processing program causing a computer to execute a function of determining whether a pixel in the image is a chromatic color or an achromatic color by comparing the difference value with a predetermined threshold value.

Images