JP4748041B2 - Image processing method, program, and image processing apparatus - Google Patents

Description

  The present invention relates to an image processing method, a program, and an image processing apparatus for processing a color image, and more particularly to an image processing technique for reducing a color difference between pixels constituting a color image.

  In the manufacturing field, automation is being promoted from the viewpoint of labor saving and high efficiency. In order to realize automation, many sensors using light, electricity, radio waves, sound waves and the like are used. Among such sensors, an image processing technique that can photograph a product, a semi-finished product, and the like and process the photographed image to determine whether the product is good or not and to manage production of the product is often used. ing. According to the image processing technology, since the detection function similar to the detection by human vision can be realized, the application range is wide.

  In such an image processing technique, defect detection processing for detecting dirt or scratches to be detected is frequently performed. Such a defect detection process is realized based on a color difference between pixels constituting a captured image. For example, the presence / absence of a defective portion is determined based on whether or not the color difference between adjacent or adjacent pixels exceeds a predetermined threshold value.

  By the way, when the detection target is a single color, the color difference between the single color and the color of the defective portion is relatively easily calculated, so that the problem of erroneous detection hardly occurs. On the other hand, when a logo mark or the like is printed on the detection target, an image including three colors of the background color of the detection target, the color of the logo mark, and the color of the defective portion is obtained. And the color of the logo mark may be erroneously determined to be a defective part. A similar phenomenon can also occur due to variations in illumination during shooting.

  As described above, the “color unevenness” generated when the detection target is photographed reduces the detection accuracy of the defective portion. For this reason, it is necessary to perform preprocessing that attenuates the color unevenness in the image and smoothes it as much as possible.

As a conventional color correction technique, for example, in Japanese Patent Laid-Open No. 5-28236 (Patent Document 1), only a thin portion of a ground color is reproduced in white, and a light chromatic color of an image portion can be reproduced as a document image. A color image processing apparatus configured as described above is disclosed. According to this color image processing apparatus, there is provided means for converting the color image signal into a specific color signal when all of the plurality of color components have values within a predetermined range. Japanese Laid-Open Patent Publication No. 7-147629 (Patent Document 2) discloses a color image processing apparatus capable of replacing a specific color portion in an input image with a desired color designated with high accuracy. Yes. According to this color image processing apparatus, a determination unit that determines whether or not a color image signal before correction by the correction unit includes a specific color, and a conversion that converts the specific color determined by the determination unit into another specific color. Means.
Japanese Patent Laid-Open No. 5-28236 JP-A-7-147629

  However, the color image processing apparatus disclosed in the above-mentioned Japanese Patent Application Laid-Open No. 5-28236 (Patent Document 1) and Japanese Patent Application Laid-Open No. 7-147629 (Patent Document 2) improves the color reproducibility of the original color image. It was intended to be made. For this reason, if color unevenness exists in a color image, the color unevenness is to be reproduced more faithfully, and a process for attenuating and smoothing the color unevenness has not been assumed. Therefore, it has not been possible to realize the pre-processing of the defect detection process as described above using a conventional color correction technique.

  Accordingly, the present invention has been made to solve such a problem, and an object of the present invention is to provide an image processing method, a program, and an image processing apparatus capable of smoothing color unevenness in a color image. is there.

  According to an aspect of the present invention, there is provided an image processing method for reducing a color difference between pixels with respect to a color image composed of a plurality of pixels. The image processing method according to this aspect includes a step of defining a reference line including two reference points on the color coordinates corresponding to the first and second colors to be selected on the color coordinates, and a predetermined processing of the color image An extraction step for sequentially extracting each pixel included in the region, and a conversion line on the color coordinate including the conversion source point on the color coordinate corresponding to the color of the pixel extracted in the extraction step is defined on the color coordinate A conversion step for calculating a post-conversion point obtained by moving the conversion source point by a predetermined movement amount according to the conversion line specified in the specification step, and calculating a color of the extracted pixel in the conversion step A replacement step for replacing the converted color with a color corresponding to the converted point, and a extracting step, a conversion step, and a conversion step until all pixels included in the processing region are extracted. And a step of repeatedly executing the substitution step.

  According to the invention according to this aspect, the color of the pixel to be converted is changed in accordance with the conversion line defined along the reference line including the two reference points on the color coordinates corresponding to the selected first and second colors. A post-conversion point is calculated by moving the corresponding conversion source point by a predetermined amount of movement. Then, the color of the pixel to be converted is replaced with a color corresponding to the post-conversion point. In this way, each pixel included in the predetermined processing region of the color image is replaced with a color calculated according to the conversion line defined along the reference line, so that unified color correction as a whole is performed. Done. Thereby, the color nonuniformity in a color image can be smoothed.

Preferably, the reference line is defined as a linear expression on the color coordinates.
More preferably, the conversion line is defined to be parallel to the reference line on the color coordinates.

  Preferably, the image processing method according to this aspect further includes a step of determining a movement amount with respect to the conversion source point based on a distance from the conversion source point to the reference line.

  Preferably, in the conversion step, the post-conversion point is obtained by moving the conversion source point in a direction approaching a predetermined convergence point existing on the reference line.

More preferably, the convergence point is the midpoint between the two reference points.
Preferably, the convergence point coincides with one of the two reference points.

  According to another aspect of the present invention, an image processing method for reducing a color difference between pixels in a color image composed of a plurality of pixels. The image processing method according to this aspect includes a step of defining a reference line including two reference points on the color coordinates corresponding to the first and second colors to be selected on the color coordinates, and a predetermined processing of the color image An extraction step for sequentially extracting each pixel included in the region, a defining step for defining on the color coordinate a conversion line including a conversion source point on the color coordinate corresponding to the color of the pixel extracted in the extraction step, and conversion Based on the distance from the original point to the reference line defined in the defining step, a determining step for determining the amount of movement with respect to the converting source point and the converting source point determined in the determining step according to the conversion line defined in the defining step A conversion step for calculating a post-conversion point obtained by moving the amount of movement, and a color corresponding to the post-conversion point calculated in the conversion step. Comprising a substitution step of conversion, it extracted until all the pixels contained in the processing region is extracted in step, defining step, determination step, and a step of repeatedly executing the conversion step and replacing step.

  According to the invention according to this aspect, based on the distance to the reference line including two reference points on the color coordinates corresponding to the first and second colors selected from the conversion source point corresponding to the color of the pixel to be converted. To determine the amount of movement for obtaining the converted point. Then, the converted point is calculated by moving the conversion source point by the movement amount according to the conversion line. On the color coordinates, the distance from the conversion source point to the reference line indicates the closeness between the color of the pixel to be converted and the first color or the second color, and thus approximates the first color or the second color. While correcting the larger color, the color correction of the other colors (for example, the color of the defective portion described above) can be suppressed. As a result, the color unevenness in the background region can be smoothed while maintaining the color of the defective portion to be detected in the color image.

  Preferably, in the determination step, the movement amount is determined to be a larger value as the distance from the conversion source point to the reference line becomes smaller.

  Preferably, the above-described image processing method represents the color extraction region based on the step of accepting the designation of the color extraction region including at least a part of the color image and the color of each pixel included in the color extraction region. Selecting two colors as the first and second colors.

Preferably, in the image processing method described above, when the third color is selected in addition to the first and second colors, the series of steps is performed after the series of steps is performed based on the first and second colors. The method further includes a step of executing the series of steps again based on the color corresponding to the convergence point generated by the execution of the step and the third color. Preferably, the movement amount is a variable from the distance from the conversion source point to the reference line. The function is determined based on the function, and the function is configured such that the relationship between the distance and the movement amount is variable according to a parameter set from the outside.

  Preferably, the image processing method further includes a step of simultaneously displaying a color image and at least one of the visual display and the color information of the first and second colors on the same screen.

Preferably, the color coordinates include three coordinate axes corresponding to the three primary colors.
According to still another aspect of the present invention, there is provided a program for causing a computer to execute an image processing method for reducing a color difference between pixels for a color image composed of a plurality of pixels. The program according to this aspect includes a step of defining a reference line including two reference points on the color coordinates corresponding to the first and second colors to be selected on the color coordinates, and a predetermined processing area of the color image. An extraction step for sequentially extracting each pixel included in the image, and a rule that includes a conversion source point on the color coordinate corresponding to the color of the pixel extracted in the extraction step and defines a conversion line on the color coordinate along the reference line A conversion step for calculating a post-conversion point obtained by moving the conversion source point by a predetermined movement amount in accordance with the conversion line defined in the defining step, and a color of the extracted pixel calculated in the converting step A replacement step for replacing with a color corresponding to the post-conversion point, and until all the pixels included in the processing region are extracted in the extraction step, the defining step, the conversion step, and And a step of repeatedly executing the conversion step.

  According to still another aspect of the present invention, there is provided a program for causing a computer to execute an image processing method for reducing a color difference between pixels for a color image composed of a plurality of pixels. The program according to this aspect includes a step of defining a reference line including two reference points on the color coordinates corresponding to the first and second colors to be selected on the color coordinates, and a predetermined processing area of the color image. An extraction step for sequentially extracting each pixel included in the image, a defining step for defining a conversion line on the color coordinate on the color coordinate corresponding to the color of the pixel extracted in the extraction step, and a conversion source point A determination step for determining a movement amount with respect to the conversion source point based on a distance from the reference line specified in the specification step to the reference line, and a movement determined in the determination step for the conversion source point according to the conversion line specified in the specification step A conversion step for calculating a post-conversion point obtained by moving the image by an amount, and setting the color of the extracted pixel to a color corresponding to the post-conversion point calculated in the conversion step. Comprising a substitution step of the extraction until all the pixels contained in the processing region is extracted in step, defining step determination step, and a step of repeatedly executing the conversion step and replacing step.

  According to still another aspect of the present invention, an image processing apparatus for reducing a color difference between pixels in a color image composed of a plurality of pixels. An image processing apparatus according to this aspect includes a first defining unit that defines, on a color coordinate, a reference line that includes two reference points on the color coordinate corresponding to the first and second colors to be selected, and a predetermined color image. Extraction means for sequentially extracting each pixel included in the processed area, and a conversion line that includes a conversion source point on the color coordinate corresponding to the color of the pixel extracted by the extraction means, and that is along the reference line Second defining means defined above, conversion means for calculating a converted point obtained by moving the conversion source point by a predetermined movement amount according to the conversion line defined by the second defining means, and the extracted pixels Replacement means for substituting the color corresponding to the post-conversion point calculated by the conversion means.

  According to still another aspect of the present invention, an image processing apparatus for reducing a color difference between pixels in a color image composed of a plurality of pixels. An image processing apparatus according to this aspect includes a first defining unit that defines, on a color coordinate, a reference line that includes two reference points on the color coordinate corresponding to the first and second colors to be selected, and a predetermined color image. Extraction means for sequentially extracting each pixel included in the processed area, and a conversion line including a conversion source point on the color coordinate corresponding to the color of the pixel extracted by the extraction means is defined on the color coordinate. Based on the defining means, the determining means for determining the movement amount with respect to the conversion source point based on the distance from the conversion source point to the reference line defined by the second defining means, and the conversion source according to the conversion line defined by the defining means A conversion unit that calculates a post-conversion point obtained by moving the point by the amount of movement determined by the determination unit, and a replacement that replaces the color of the extracted pixel with a color corresponding to the post-conversion point calculated by the conversion unit With means .

  According to the present invention, it is possible to realize an image processing method, a program, and an image processing apparatus that can smooth color unevenness in a color image.

  Embodiments of the present invention will be described in detail with reference to the drawings. Note that the same or corresponding parts in the drawings are denoted by the same reference numerals and description thereof will not be repeated.

  FIG. 1 is a schematic configuration diagram of an image processing system 100 including an image processing apparatus 1 according to an embodiment of the present invention.

  Referring to FIG. 1, an image processing system 100 includes an image processing device 1, an imaging unit 2, and a display unit 3. As an example, an imaging unit detects a detection target that is continuously conveyed on a production line or the like. The image processing apparatus 1 performs color unevenness smoothing processing (hereinafter, also referred to as “smoothing processing”) according to the present embodiment on the captured color image. Then, the image processing apparatus 1 performs defect detection processing on the color image that has been subjected to smoothing processing, which is preprocessing, and determines whether or not a defective portion such as a stain or a scratch exists in the detection target. To do. Further, the image processing device 1 displays the detection result on the display unit 3 and outputs the detection result to another device (not shown). The detection target is, for example, a sheet-like object such as paper or film, or a metal product.

  As an example, the imaging unit 2 includes an imaging element and a lens such as a CCD (Coupled Charged Device) or a CMOS (Complementary Metal Oxide Semiconductor) sensor, captures a detection target, and outputs the captured color image to the image processing apparatus 1. To do. Note that the color image captured by the imaging unit 2 may be either a still image or a moving image. The color image captured by the imaging unit 2 includes a plurality of pixels arranged in a matrix, and each pixel has color information. As an example of color information, in the present embodiment, RGB information including gray values of “red”, “green”, and “blue” based on the three primary colors is used. Instead of RGB information, CMY information consisting of shades of “cyan”, “magenta”, and “yellow”, which are complementary colors of the three primary colors, “Hue”, “Value”, and “Chroma” ) "May be used as the color information.

  The display unit 3 displays a processing result in the image processing device 1, an image taken by the imaging unit 2, a user interface screen for operating the image processing device 1, and the like. As an example, the display unit 3 includes a liquid crystal display (LCD), a plasma display, an EL display (Electro Luminescence display), and the like.

  The image processing apparatus 1 includes a CPU unit 4, an auxiliary storage unit 5, an input unit 6, an imaging unit interface (imaging unit I / F) 7, a main storage unit 8, a display processing unit 9, and an external interface ( The external I / F 10, the reading unit 11, and the bus 13 are realized by a personal computer as an example.

  The imaging unit interface 7 is electrically connected to the imaging unit 2, receives a video signal captured by the imaging unit 2, performs predetermined signal conversion processing, acquires color information of each pixel, and then transmits the color information to the bus. 13 to the CPU unit 4 via Specifically, the imaging unit interface 7 performs frame synchronization with the video signal received from the imaging unit 2, demodulates the color information of each pixel that is expanded and transmitted on the time axis, and The red, blue, and green shade values (hereinafter also simply referred to as “RGB information”) are acquired.

  The main storage unit 8 stores a program executed by the CPU unit 4, a color image taken by the imaging unit 2, a color image during image processing in the CPU unit 4, and the like. For example, the main storage unit 8 includes a semiconductor storage element such as a DRAM (Dynamic Random Access Memory) or an SRAM (Static Random Access Memory).

  The display processing unit 9 receives display data for displaying an image after processing in the CPU unit 4, an image taken by the imaging unit 2, a user interface screen, a processing state in the CPU unit 4, and performs predetermined signal processing. After performing, it outputs to the display part 3 as a video signal.

  The external interface 10 outputs a processing result executed by the CPU unit 4 to the outside. As an example, the external interface 10 includes a contact output (DO) composed of a photodiode, a transistor, or a relay, USB (Universal Serial Bus), RS-232C (Recommended Standard 232 version C), IEEE (Institute of Electrical and It consists of communication means in accordance with Electronic Engineers (1394), Small Computer System Interface (SCSI), Ethernet (registered trademark), and the like.

  The auxiliary storage unit 5 includes a non-volatile storage area, and stores images captured by the imaging unit 2, images processed by the CPU unit 4, processing results, and the like. As an example, the auxiliary storage unit 5 includes a semiconductor memory such as a hard disk drive (HDD), a flash memory card, an SD memory card, and an IC memory card.

  The input unit 6 receives settings and instructions from the user and supplies them to the CPU unit 4 via the bus 13. As an example, the input unit 6 includes a keyboard, a mouse, and a touch panel or a touch pen provided on the display surface of the display unit 3.

  The reading unit 11 receives the recording medium 12 on which the program executed by the CPU unit 4 is recorded, reads the program, and supplies the program to the auxiliary storage unit 5 or the main storage unit 8. The recording medium 12 may be any medium that holds data in a nonvolatile manner. As an example, the recording medium 12 is an optical disc (CD (Compact Disk), DVD (Digital Versatile Disc) -ROM / RAM / R / RW, MO (Magnetic Optical). Disc), MD (Mini Disc)), a flexible disk, and a removable recording medium such as a magnetic tape.

  The CPU unit 4 receives the RGB information generated from the color image photographed by the imaging unit 2 via the imaging unit interface 7 and stores it in the main storage unit 8 in association with the coordinates of each pixel. Then, the CPU unit 4 sequentially executes a smoothing process, which will be described later, on each pixel included in the processing region in accordance with the setting in the processing region given in advance via the input unit 6, and after the replacement Are stored in the main storage unit 8. In addition, the CPU unit 4 outputs the color image after the smoothing process to the display processing unit 9 and the like, and executes a defect detection process and the like on the color image after the smoothing process.

(Outline of smoothing process)
FIG. 2 is a schematic diagram for illustrating an outline of the smoothing process according to the embodiment of the present invention.

FIG. 2A is an example of a color image before the smoothing process.
FIG. 2B is an example of a color image after the smoothing process.

  With reference to FIG. 2A, a case where the detection target OBJ has a logo mark 30 composed of characters “PATENT” on the ground color (for example, white) is shown. The logo mark 30 has a predetermined color different from the ground color. The smoothing process according to the present embodiment is intended for the case of detecting a defective portion 40 that occurs in such a detection target OBJ.

  As described above, when the color difference between adjacent pixels in such a color image is calculated, a large value is obtained at the boundary between the ground color and the logo mark 30 and the boundary between the ground color and the defective portion 40. For this reason, the boundary between the ground color and the logo mark 30 is erroneously detected as a defective portion.

  Therefore, as shown in FIG. 2B, it is necessary to smooth the color unevenness (color difference) including the ground color of the detection target OBJ and the logo mark 30 so that only the defective portion 40 is “relieved”. As an example of such smoothing, in FIG. 2B, the ground color of the detection target OBJ and the logo mark 30 are brought close to an intermediate color, that is, a specific color existing between the ground color and the logo mark 30 color. As described above, color correction is performed.

  As described above, the smoothing process refers to the color originally included in the detection target OBJ, that is, the color of the area excluding the defective part 40 in the color image (hereinafter also referred to as “background area”) between the defective part 40. It means correction that approaches a specific color that can ensure a relatively large color difference.

  The background area of the color image of the detection target OBJ that is actually captured is not necessarily composed of simple two colors as conceptually shown in FIG. That is, the imaging unit 2 (FIG. 1) discriminates a subtle difference in color as the color resolution is higher, and outputs it as pixels having different colors. A similar phenomenon can also occur due to a difference in illumination light irradiation during photographing. As a result, the captured color image includes a plurality of pixels having color information that is similar but not identical. Therefore, even if only two specific colors (first color and second color) are replaced with a specific intermediate color, sufficient smoothing processing cannot be realized. Therefore, in the present embodiment, conversion processing on color coordinates is performed as described below.

FIG. 3 is a diagram for illustrating the contents of the smoothing process according to the embodiment of the present invention.
Referring to FIG. 3, in the present embodiment, conversion processing is performed on color coordinates SYS in order to realize smoothing processing. The color coordinate is a coordinate system in which parameters included in the color information of each pixel are defined in association with coordinate axes. In the present embodiment, RGB information is used as an example, so that three-dimensional color coordinates (color coordinate space) defined by the R axis, the G axis, and the B axis can be defined. The color information of each pixel constituting the color image is uniquely associated with such coordinate points on the color coordinate SYS. Although the color coordinates can be expanded to two dimensions or four dimensions or more according to the type of color information parameter, the processing described below can be applied to any number of color coordinates.

  In the present embodiment, the reference line CDV including the reference points COL1 and COL2 corresponding to the selected first color and second color is defined on the color coordinate SYS, and each reference line CDV is used to determine each reference line CDV. Pixel color is corrected.

  Here, the user can arbitrarily select the first color and the second color, but in many cases, the user can select one color corresponding to the ground color of the detection target OBJ as shown in FIG. One color corresponding to the color of the logo mark 30 is selected. In the following description, it is assumed that the first color selected by the user has R1, G1, and B1 as color information values, and the second color has R2, G2, and B2 as color information values, respectively.

  When the first color and the second color are selected in this way, the reference line CDV including the reference points COL1 and COL2 on the color coordinates SYS corresponding to these is defined. In the present embodiment, the reference line CDV is defined as a linear expression on the color coordinates SYS, that is, defined as a straight line passing through the reference points COL1 and COL2. Further, a reference plane CPL orthogonal to the convergence point CNT located at the midpoint between the reference point COL1 and the reference point COL2 on the reference line CDV is defined.

  Subsequently, pixels included in the color image processing region are sequentially extracted. In response to this pixel extraction, a conversion line #CDV including the conversion source point PEL on the color coordinate SYS corresponding to the color of the pixel and along the reference line CDV is defined. In the present embodiment, since the reference line CDV is defined as a straight line, the conversion line #CDV is also defined as a straight line that is parallel to the reference line CDV and includes the conversion source point PEL.

  Subsequently, a post-conversion point #PEL obtained by moving the conversion source point PEL by a predetermined amount of movement in a direction approaching the convergence point CNT according to the conversion line #CDV is calculated. Then, the color of the extracted pixel is replaced with a color corresponding to the calculated post-conversion point #PEL.

  The above processing is sequentially executed for all the pixels included in the color image processing region.

  Conceptually speaking, the color of each pixel included in the color image to be processed is corrected so as to be directed to the reference plane CPL on the conversion line #CDV. Therefore, each pixel is corrected so as to approach the color corresponding to the intersection of the reference plane CPL and the conversion line #CDV.

  Here, it is also important how close the color of each pixel is to the reference plane CPL. That is, in order to execute a defect detection process or the like based on the color difference, it is necessary to keep the color corresponding to the defect portion 40 (FIG. 2) as close as possible to the reference plane CPL. In other words, a pixel having a color close to the background color of the detection target OBJ and the logo mark 30 as shown in FIG. 2 needs to be as close to the reference plane CPL as possible. Needs to be as far as possible from the reference plane CPL.

  Therefore, in the present embodiment, the movement amount with respect to the conversion source point PEL is determined based on the distance w from the conversion source point PEL to the reference line CDV. More specifically, the movement amount is determined so as to increase as the distance w decreases. In other words, the smaller the color difference between the color range corresponding to the reference line CDV including the reference points COL1 and COL2, the more the correction is made so that it is closer to the reference plane CPL.

  By determining the amount of movement in this way, the pixels close to the selected first color and second color are corrected (smoothed) more greatly, but separated from the first color and the second color. Pixels corresponding to the defective portion 40 and the like are hardly corrected. As a result, the background region of the color image is focused and smoothed, and the defective portion can be detected more accurately.

(Smoothing process)
FIG. 4 is a diagram for describing the process of the smoothing process according to the embodiment of the present invention. 4, reference points COL1, COL2, a reference line CDV, and the like are shown with reference to the reference plane CPL shown in FIG.

  FIG. 5 is a flowchart showing an outline of the processing procedure of the smoothing processing according to the embodiment of the present invention. The following processing procedure is realized when the CPU unit 4 (FIG. 1) of the image processing apparatus 1 executes a program.

  Referring to FIGS. 4 and 5, CPU unit 4 accepts the first color and the second color selected by the user by operating input unit 6 (step S2). Note that the user can easily select the first color and the second color according to a color extraction process described later. Then, the CPU unit 4 defines a reference line CDV including reference points COL1 and COL2 corresponding to the first color and the second color on the color coordinate SYS (step S4). Furthermore, the CPU unit 4 defines a reference plane CPL that is orthogonal to the midpoint (convergence point CNT) between the reference point COL1 and the reference point COL2 on the reference line CDV (step S6).

  The CPU unit 4 extracts pixels included in the color image processing area (step S8). The processing area can be arbitrarily set by the user, and the whole or a part of the color image photographed by the imaging unit 2 (FIG. 1) is set as the target area for the smoothing process. Then, the CPU unit 4 defines the conversion line #CDV including the conversion source point PEL on the color coordinate SYS corresponding to the extracted pixel, and calculates the height h from the conversion source point PEL to the reference plane CPL ( Step S10).

  Further, the CPU unit 4 calculates the distance w from the conversion source point PEL to the conversion line CDV (step S12). Then, the CPU unit 4 calculates a post-conversion point #PEL obtained by moving the predetermined amount of movement along the conversion line #CDV based on the distance w calculated in step S12 (step S14). More specifically, the CPU unit 4 calculates the height #h from the post-conversion point #PEL to the reference plane CPL. Therefore, (height h−height #h) is the amount of movement.

  Further, the CPU unit 4 replaces the color of the pixel extracted in step S8 with a color corresponding to the post-conversion point #PEL (step S16).

  Thereafter, the CPU unit 4 determines whether or not all the pixels included in the color image processing area have been extracted (step S18). If all the pixels have not been extracted (NO in step S18), the CPU unit 4 repeatedly executes the processes from step S8 to S18 described above.

  When all the pixels have been extracted (YES in step S18), the CPU unit 4 displays the color image including the pixels after color replacement, the color display of the first color and the second color, and RGB Information is displayed on the display unit 3 (step S20), and the smoothing process is terminated.

  Hereinafter, the contents of the arithmetic processing in the main steps shown in FIG. 5 will be described with reference to FIG.

(1) Definition of reference line CDV (step S4)
In order to define the reference line CDV, a vector CDV from the reference point COL1 to the reference point COL2 on the color coordinate SYS (in FIG. 4, an arrow is added to the upper part of the character to indicate the vector) is calculated. Here, each component of the vector CDV is
Vector CDV = (VR, VG, VB) = (R2-R1, G2-G1, B2-B1)
It is. Further, the components VRn, VGn, VBn when the vector CDV is normalized, that is, the norm (magnitude) thereof is set to “1”,
(VRn, VGn, VBn) = (VR / NRM, VG / NRM, VB / NRM)
However, NRM ² = VR ² + VG ² + VB ²
And

(2) Definition of reference plane CPL (step S6)
In the present embodiment, since the convergence point CNT is defined at the midpoint between the reference point COL1 and the reference point COL2, a reference plane CPL passing through the convergence point CNT and orthogonal to the vector CDV is defined. The coordinates of the convergence point CNT are
Convergence point CNT = (Rc, Gc, Bc) = ((R2 + R1) / 2, (G2 + G1) / 2, (B2 + B1) / 2)
It becomes.

(3) Definition of conversion line #CDV and calculation of height h from conversion source point PEL to reference plane CPL (step S10)
Since conversion line #CDV is parallel to conversion line CDV, each component of vector #CDV is the same as each component of vector CDV. That is, the conversion line #CDV passes through the conversion source point PEL and has the same vector component as the vector CDV.

The height h from the conversion source point PEL to the reference plane CPL includes the coordinate position (Rs, Gs, Bs) of the conversion source point PEL, the coordinate position (Rc, Gc, Bc) of the convergence point CNT, and the component value of the vector CDV ( VR, VG, VB)
Height h = VR × Rs + VG × Gs + VB × Bs−VR × Rc−VG × Gc−VB × Bc
Is calculated according to

(4) Calculation of distance w from conversion source point PEL to conversion line CDV (step S12)
The distance w from the conversion source point PEL to the conversion line CDV is equal to the distance between the mapping point PRJ, which is the intersection of the conversion line #CDV and the reference plane CPL, and the convergence point CNT.

Therefore, the coordinates of the mapping point PRJ are obtained by using a component value obtained by normalizing the vector CDV.
Mapping point PRJ = (Rs−h × VRn, Gs−h × VGn, Bs−h × VBn)
It becomes.

The distance (corresponding to the distance w) between the convergence point CNT and the mapping point PRJ on the reference plane CPL is:
Distance w = √ {(Rs−h × VRn−Rc) ² + (Gs−h × VGn−Gc) ² + (Bs−h × VBn−Bc) ² }
Is calculated according to

(5) Calculation of post-conversion point #PEL (step S14)
Based on the distance w calculated in the above (4), the height #h from the post-conversion point #PEL to the reference plane CPL is calculated. More specifically, it is calculated according to # h = h × f (w) using a function f (w) (where 0 ≦ f (w) ≦ 1) with the distance w as a variable.

  Here, the function f (w) has a characteristic that the movement amount of the conversion source point PEL increases as the distance w decreases, that is, the height #h decreases as the distance w decreases.

FIG. 6 is a diagram illustrating an example of the characteristics of the function f (w).
With reference to FIG. 6, any function may be used as long as it is a function f (w) having the above-described characteristics. In the present embodiment,
f (w) = w ² / (w ² + Lv ² )
Is used. The parameter Lv is configured to be settable by the user, and the characteristic of the function f (w) changes according to the parameter Lv.

  The function f (w) shown in FIG. 6 is a monotonically increasing function with respect to the distance w, and f (w) = 0.5 when the distance w matches the parameter Lv. That is, since the distance w for reducing the height h to 50% changes according to the parameter Lv, the parameter Lv indicates a “range” indicating how much the color range approaches the reference plane CPL. Will be defined.

Referring to FIG. 4 again, using the height #h calculated in the above-described process, the coordinates of the converted point #PEL are the coordinates of the mapping point PRJ (Rs−h × VRn, Gs−h × VGn). , Bs−h × VBn)
Post-conversion point # PEL = (Rs + (# h−h) × VRn, Gs + (# h−h) × VGn, Bs + (# h−h) × VBn)
It becomes.

  Since each component of the coordinates of the post-conversion point #PEL corresponds to the shade value of “red”, “green”, and “blue”, each component value is replaced as color information of the extracted pixel.

  On the conversion line #CDV, the point closest to the convergence point CNT (the point at which the distance to the convergence point CNT is the shortest) is the mapping point PRJ. Therefore, in the present embodiment, the “direction toward the convergence point CNT” Means “direction on conversion line #CDV toward mapping point PRJ”.

(Example of screen display in smoothing process)
With reference to an example of the screen display in display unit 3 according to the present embodiment, the smoothing processing execution operation according to the present embodiment will be further described.

FIG. 7 is a diagram illustrating a screen display example immediately after a color image of the detection target OBJ is captured.
FIG. 8 is a diagram showing a screen display example when selecting the first color and the second color.

FIG. 9 is a diagram illustrating a screen display example after the smoothing process is executed.
FIG. 10 is a diagram illustrating a screen display example after executing the smoothing process by setting the value of the parameter Lv to be larger than that in FIG. 9.

  Referring to FIG. 7, in display unit 3, the captured color image of detection object OBJ is displayed on the right side, and for the visual display of the first color and the second color and the display / setting of color information. The area is displayed on the left.

  The user can select and set by operating the input unit 6 and moving the pointer 50 on the screen. First, the user selects a desired color image from among color images photographed in real time by the imaging unit 2 or color images photographed in advance, and displays the selected color image on the display unit 3.

  As an example, FIG. 7 shows a case where a defective portion 40 is generated with respect to a checkered pattern detection target OBJ. Furthermore, it is assumed that there is a radial shading difference in the luminance of the detection target OBJ.

  When the color image is displayed, the user operates the pointer 50 to set the processing area AREA. This processing area AREA is set by the upper left start point START and the lower right end point END. In FIG. 7, in order to cut an area corresponding to the production line, a processing area AREA is set corresponding to the detection target OBJ. Also, the coordinates of the start point START in the image are (Xs, Ys), and the coordinates of the end point END are (Xe, Ye).

  In addition, a first color display frame 54 and a second color display frame 64 are provided in the left area of the display unit 3, and RGB display frames 56 and 66 for displaying the respective color information (RGB information as an example). Is provided. Further, parameter setting dials 58 and 68 are provided in association with the RGB display frames 56 and 66, respectively, so that the user can arbitrarily set the color information of the first color and the second color.

  When the setting of the processing area AREA is completed, the user presses the “image designation” button 52 in order to select the first color and the second color. Then, the screen is switched to the screen shown in FIG.

  The user operates the pointer 50 to specify the color extraction region 42 so as to include color unevenness (checkered pattern) in the color image. When the specification of the color extraction area 42 is completed, the user presses the “OK” button 70. Then, according to the color extraction process described later, two colors representing the color extraction area 42 are selected as the first color and the second color from the colors of the respective pixels included in the color extraction area 42. Then, the above-described smoothing process is executed according to the selected first color and second color. The color image after the smoothing process is executed is displayed as shown in FIG.

  As shown in FIG. 9, the background area excluding the defective portion 40 is substantially smoothed by the smoothing process according to the present embodiment. FIG. 9 shows an example in which the area other than the processing area AREA set by the user in FIG. 7 is masked and only the color image in the processing area AREA is displayed.

  In FIG. 9, the selected first color and second color are visually displayed in the first color display frame 54 and the second color display frame 64, and in the RGB display frames 56 and 66, The RGB information of the first color and the second color is displayed as numerical values. Note that the user can further change the selected first color and second color by operating the parameter setting dials 58 and 68.

  In response to the user's operation of the parameter setting dials 58 and 68, the smoothing process is executed again. Note that the original color image and the converted color image are independently stored in the image processing apparatus 1 so that the smoothing process can be re-executed.

  Here, in FIG. 9, color unevenness remains in the area on the right side of the processing area AREA due to the radial density difference occurring in the detection target OBJ. That is, the smoothing process is insufficient for the color image of the detection target OBJ illustrated.

  Therefore, referring to FIG. 10, when the user changes the value of parameter Lv set in “color unevenness cut level” setting frame 72 provided at the lower left of display unit 3 by operating parameter setting dial 74, the change is made. The smoothing process is re-executed based on the parameter Lv. Then, the color image obtained by the re-execution is displayed on the display unit 3. In this way, the user can also set the optimum parameter Lv while viewing the smoothed color image displayed on the display unit 3.

(Color extraction processing)
As described with reference to FIG. 8, two colors representing the color extraction region 42 are selected based on the color of each pixel included in the designated color extraction region 42. Specifically, each pixel included in the color extraction region 42 is classified into two clusters by a clustering method based on the color information, and a color representing each classified cluster is determined.

FIG. 11 is a diagram for explaining the outline of the color extraction processing.
FIG. 11A shows an initial state of the clustering process.

FIG. 11B shows a completed state of the clustering process.
With reference to FIG. 11A, the case where eight pixels PEL1 to PEL8 are included in the color extraction region 42 will be described as an example.

First, each pixel included in the color extraction region 42 is extracted, and the centroids V1 and V2 of the two clusters are determined based on the color information of the pixel, that is, RGB information. In the present embodiment, the centroids V1 and V2 are determined as the maximum value and the minimum value of the R value, G value, and B value of each pixel. That is, if the color information of the pixel PELi (1 ≦ i ≦ 8) is (Ri, Gi, Bi,),
V1 = (Rmax, Gmax, Bmax)
V2 = (Rmin, Gmin, Bmin)
However,
Rmax = max (Ri), Gmax = max (Gi), Bmax = max (Bi)
Rmin = min (Ri), Gmin = min (Gi), Bmin = min (Bi)
It becomes. The center of gravity V1 determined in this way is obtained by independently extracting the maximum value from the RGB information of the extracted pixels, and the center of gravity V2 is obtained by independently extracting the minimum value from each other. It is.

  Subsequently, the distance between each of the pixels PEL1 to PEL8 and the centroids V1 and V2 is sequentially calculated. Among the pixels PEL1 to PEL8, those closer to the center of gravity V1 are classified as cluster 1, and those closer to the center of gravity V2 are classified into cluster 2.

  Thus, after each pixel is initially classified, new centroids # V1 and # V2 of each cluster are calculated from the average value of the RGB information of a plurality of pixels belonging to each cluster. Further, the distance between each of the pixels PEL1 to PEL8 and the calculated new centroids # V1 and # V2 is sequentially calculated, and the cluster classification is performed again according to the same procedure as described above.

  Thereafter, the same processing is repeated until the pixel to which the pixel belongs is not changed by the cluster classification again or until a predetermined number of times is reached. Then, as shown in FIG. 11B, the pixels PEL1 to PEL4 are classified so as to belong to the cluster 1, and the pixels PEL5 to PEL8 belong to the cluster 2. Then, two colors corresponding to the centroids # V1 and # V2 at this time are selected as the first color and the second color.

  Note that a color (for example, black: (R, G, B) that causes a large color difference with other pixels in the designated color extraction region 42 due to defects or noise of the image sensor that constitutes the imaging unit 2. ) = (0,0,0)) may be included. When the center of gravity # V1 or # V2 is calculated using such a noise pixel, a color different from the original two colors representing the pixels included in the color extraction region 42 may be selected. Therefore, before the first color and the second color are selected, it is determined whether or not the clustering process has been normally completed based on the number of pixels included in each cluster.

  Specifically, when the above-described series of clustering is completed, the number of pixels belonging to each cluster is calculated, and the ratio of the total number of pixels to the normal range (for example, 10% or more of the total number of pixels is included in each cluster). Is determined). If the ratio of the number of pixels assigned to any cluster is outside the normal range, it is determined that the clustering process has failed. Further, the clustering process is executed again except for the pixels assigned to the cluster outside the normal range. In this way, the influence of noise pixels is avoided.

  FIG. 12 is a flowchart illustrating the processing procedure of the color extraction processing. The following processing procedure is realized when the CPU unit 4 (FIG. 1) of the image processing apparatus 1 executes a program.

  Referring to FIG. 12, CPU unit 4 accepts designation of color extraction region 42 for a color image (step S100). Then, the CPU unit 4 extracts RGB information of pixels included in the designated color extraction area 42 (step S102).

  The CPU unit 4 calculates initial centroids V1 and V2 based on the extracted RGB information (step S104). Then, the CPU unit 4 sequentially calculates the distance between each pixel and the centroids V1 and V2 (step S106), and classifies the ones close to the centroid V1 or V2 as clusters 1 or 2 based on the calculated distances, respectively. (Step S108).

  Subsequently, the CPU unit 4 calculates new centroids # V1 and # V2 of each cluster from the average value of the RGB information of the pixels belonging to each classified cluster (step S110). Then, the CPU 4 sequentially calculates the distance between each of the pixels and the new centroids # V1 and # V2 (step S112). Further, based on the calculated distance, the CPU unit 4 reclassifies the one close to the new center of gravity # V1 or # V2 into the cluster 1 or 2 (step S114).

  Then, it is determined whether or not the pixels belonging to the reclassified clusters 1 and 2 have changed since the previous classification (step S116). If there has been a change since the previous classification (YES in step S116), the CPU unit 4 determines whether or not the cluster reclassification has been executed a predetermined number of times (step S118).

  When the cluster reclassification has not been executed a predetermined number of times (NO in step S118), the CPU unit 4 executes the processes of steps S108 to S116 described above again.

  If no change has occurred since the previous classification (NO in step S116), or if cluster reclassification has been executed a predetermined number of times (YES in step S118), the CPU unit 4 The number of pixels belonging to 1 and 2 is calculated (step S120). Then, the CPU unit 4 determines whether or not the ratio of the number of pixels belonging to each cluster to the total number of pixels is within a normal range (step S122).

  If the ratio of the number of pixels belonging to any cluster to the total number of pixels is not in the normal range (NO in step S122), the CPU unit 4 excludes pixels belonging to the cluster that is not in the normal range (step S122). S124), the process after step S104 described above is executed again.

  When the ratio of the number of pixels belonging to each cluster with respect to the total number of pixels is within the normal range (YES in step S122), the CPU unit 4 selects two colors corresponding to the current centroids # V1 and # V2. The first color and the second color are selected (step S126). Then, the CPU unit 4 ends the color extraction process.

(First modification)
In the present embodiment described above, the case where the reference plane CPL is orthogonal at the midpoint between the reference point COL1 and the reference point COL2 on the reference line CDV is exemplified. This means that the color of each pixel is corrected toward the intermediate color between the first color and the second color, but the reference plane does not necessarily have to be orthogonal at the midpoint between the reference point COL1 and the reference point COL2, It may be orthogonal at any position.

  FIG. 13 is a diagram for describing the position of reference plane #CPL according to the first modification of the embodiment of the present invention.

  Referring to FIG. 13, as an example, when a reference plane #CPL that includes a reference point COL2 corresponding to the second color is used, each pixel is corrected to approach the reference plane #CPL. Color correction is performed so as to approach the color corresponding to COL2. By changing the position of the reference plane #CPL in this way, for example, when the second color is white or the like, the background area can be smoothed to a light color as a whole, so that correction is made toward the intermediate color. In some cases, the defective portion 40 can be detected with higher accuracy.

  Note that the user can arbitrarily set the reference plane at any position. Further, except for the point that the height h between the reference plane #CPL and the pixel PEL is increased, the moving direction of the pixel PEL is the same as that of the present embodiment described above, and therefore the detailed description will not be repeated. .

(Second modification)
In the present embodiment described above, the case where a linear reference line CDV is used, which is defined as a linear expression, is exemplified, but the present invention is not limited to this.

  FIG. 14 is a diagram for illustrating a reference line #CDV according to the second modification of the embodiment of the present invention.

Referring to FIG. 14, for example, a curved reference line CDV defined as a multi-order expression including reference points COL1 and COL2 may be used. Since the reference line CDV serves as a reference in the direction in which the color correction of each pixel is performed, by defining the reference line CDV so as to bypass the coordinate point on the color coordinate SYS corresponding to the color of the defective portion 40. The background region can be smoothed while maintaining a relatively large color difference with the color of the defect portion 40. In addition, when color attributes composed of hue, brightness, and saturation are used as color information, the weights between the elements are not the same, so that the reference line CDV reflecting the weighting coefficient is set according to the items to be emphasized. It is desirable to use it. Further, in a coordinate system that is not an orthogonal coordinate system such as the L ^* a ^* b ^* color system or the L ^* C ^* h color system, a curved reference line CDV is inevitably used.

  When such a curved reference line CDV is used, the conversion line #CDV is a curve having the same shape passing through a position away from the reference line CDV by a predetermined distance.

(Third Modification)
In the present embodiment described above, the case where the smoothing process is performed using two colors (first color and second color) as a reference is exemplified. However, three colors are used in the background region of the color image of the detection target OBJ actually captured. The above may be included. Thus, even when three or more colors are included, it can be dealt with by repeatedly executing the smoothing process according to the present embodiment described above.

  FIG. 15 is a diagram for illustrating the contents of the smoothing process according to the third modification of the embodiment of the present invention.

  With reference to FIG. 15, as an example, a case where the third color is selected in addition to the first color and the second color will be described. As in FIG. 3, the reference points COL1 and COL2 correspond to the first color and the second color, respectively, while the reference point COL3 corresponds to the third color.

First, similarly to the smoothing process according to the present embodiment described above, the first smoothing process is performed on each pixel included in the target range of the color image based on the first color and the second color. . As an example, the pixel PEL moves to the post-conversion point #PEL ⁽¹⁾ according to the conversion line #CDV ⁽¹⁾ .

The color corresponding to the convergence point CNT on the reference plane CPL ⁽¹⁾ used in the ^first smoothing process is regarded as the representative color of the color image after the smoothing process. Then, based on the color corresponding to the convergence point CNT and the third color, the second smoothing process is performed on each pixel included in the target range of the color image after the first smoothing process.

That is, the reference line ^{CDV (2)} is defined comprising a reference point COL3 corresponding to the convergence point CNT and the third color, conversion lines #CDV along the reference line ^{CDV (2) (2)} is defined. Therefore, the pixel PEL moved to the post-conversion point #PEL ⁽¹⁾ further moves to the post-conversion point #PEL ⁽²⁾ according to the conversion line #CDV ⁽²⁾ .

  According to such a procedure, the smoothing process when three colors are selected is executed. Note that the same procedure may be repeated when four or more colors are selected, and thus detailed description will not be repeated.

  According to the embodiment of the present invention, according to the conversion line #CDV defined along the reference line CDV including the two reference points on the color coordinates corresponding to the first color and the second color selected by the user. Then, a post-conversion point #PEL obtained by moving the conversion source point PEL corresponding to the color of the pixel to be converted by a predetermined movement amount is calculated. Then, the color of the pixel to be converted is replaced with a color corresponding to the post-conversion point #PEL. As described above, each pixel included in the predetermined processing area AREA of the color image is replaced with a color calculated according to the conversion line #CDV defined along the reference line CDV. Color correction can be realized. Therefore, the color unevenness in the color image can be smoothed.

  According to the embodiment of the present invention, the reference line including two reference points on the color coordinates corresponding to the first color and the second color selected from the conversion source point PEL corresponding to the color of the pixel to be converted. Based on the distance w to the CDV, the movement amount for obtaining the post-conversion point #PEL is determined. Then, the converted point #PEL is calculated by moving the conversion source point PEL by the movement amount according to the conversion line #CDV. On the color coordinate SYS, the distance from the conversion source point PEL to the reference line CDV indicates the closeness between the color of the pixel to be converted and the first color or the second color. While correcting the color approximate to the color more greatly, color correction of other colors (for example, the color of the defective portion 40) can be suppressed. As a result, the color unevenness in the background region can be smoothed while maintaining the color of the defective portion 40 to be detected in the color image.

  The embodiment disclosed this time should be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

1 is a schematic configuration diagram of an image processing system 100 including an image processing device 1 according to an embodiment of the present invention. It is a schematic diagram for demonstrating the outline of the smoothing process according to embodiment of this invention. It is a figure for demonstrating the content of the smoothing process according to embodiment of this invention. It is a figure for demonstrating the process of the smoothing process according to embodiment of this invention. It is a flowchart which shows the outline of the process sequence of the smoothing process according to embodiment of this invention. It is a figure which shows an example of the characteristic of a function f (w). It is a figure which shows the example of a screen display immediately after image | photographing the color image of the detection target OBJ. It is a figure which shows the example of a screen display when selecting a 1st color and a 2nd color. It is a figure which shows the example of a screen display after performing a smoothing process. It is a figure which shows the example of a screen display after setting the value of parameter Lv large compared with FIG. 9, and performing the smoothing process. It is a figure for demonstrating the outline of a color extraction process. It is a flowchart which shows the process sequence of a color extraction process. It is a figure for demonstrating the position of reference plane #CPL according to the 1st modification of embodiment of this invention. It is a figure for demonstrating the reference line #CDV according to the 2nd modification of embodiment of this invention. It is a figure for demonstrating the content of the smoothing process according to the 3rd modification of embodiment of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 Image processing apparatus, 2 Imaging part, 3 Display part, 4 CPU part, 5 Auxiliary storage part, 6 Input part, 7 Imaging part interface (imaging part I / F), 8 Main memory part, 9 Display processing part, 10 External Interface (external I / F), 11 reading unit, 12 recording medium, 13 bus, 30 logo mark, 40 defective part, 42 color extraction area, 50 pointer, 52 button, 54, 64 color display frame, 56, 66 RGB display Frame, 58 parameter setting dial, 70 buttons, 72 setting frame, 74 parameter setting dial, 100 image processing system, AREA processing area, OBJ detection target, SYS color coordinates.

Claims (12)

An image processing method for reducing a color difference between pixels with respect to a color image composed of a plurality of pixels,
Defining a reference line on the color coordinates including two reference points on the color coordinates corresponding to the selected first and second colors;
An extraction step of sequentially extracting each pixel included in a predetermined processing region of the color image;
A defining step including a conversion source point on the color coordinate corresponding to the color of the pixel extracted in the extraction step and defining a conversion line along the reference line on the color coordinate;
A conversion step of calculating a post-conversion point obtained by moving the conversion source point by a predetermined movement amount according to the conversion line defined in the defining step;
A replacement step of replacing the color of the extracted pixel with a color corresponding to the post-conversion point calculated in the conversion step;
Until all the pixels contained in the processing region in the extraction step is extracted, the defining step, seen including a step of repeatedly executing said conversion step and said replacing step,
The reference line is defined as a linear expression on the color coordinates,
The image processing method, wherein the conversion line is defined to be parallel to the reference line on the color coordinate . The image processing method, based on the distance from the conversion source point to the reference line, further comprising determining the amount of movement with respect to the conversion source point, the image processing method according to claim 1. The image processing method according to claim 1, wherein in the conversion step, the post-conversion point is obtained by moving the conversion source point in a direction approaching a predetermined convergence point existing on the reference line. . The image processing method according to claim 3 , wherein the convergence point is a midpoint between the two reference points. The image processing method according to claim 3 , wherein the convergence point matches one of the two reference points. The image processing method includes:
Receiving a designation of a color extraction region that includes at least a portion of the color image;
Based on the color included in each pixel included in the color extraction area, two color representative of the color extraction area further comprises the step of selecting as said first and second color, one of the claims 1-5 The image processing method according to claim 1. In the image processing method, when a third color is selected in addition to the first and second colors, a series of steps is executed based on the first and second colors, and then the series of steps is performed. The image processing method according to claim 3 , further comprising a step of executing a series of steps again based on a color corresponding to a convergence point generated by the execution and the third color. The amount of movement is determined based on a function having a variable from the conversion source point to the reference line,
The image processing method according to claim 2 , wherein the function is configured such that a relationship between the distance and the movement amount is variably set according to a parameter set from the outside. The image processing method includes the color image, and at least one of the first and second color visual displays and color information, further comprising the step of simultaneously displaying on the same screen, either Claim 1-8 The image processing method according to claim 1. The color coordinate is configured to include three coordinate axes corresponding to the three primary colors, the image processing method according to any one of claims 1-9. A program for causing a computer to execute an image processing method for reducing a color difference between pixels with respect to a color image composed of a plurality of pixels,
Defining a reference line on the color coordinates including two reference points on the color coordinates corresponding to the selected first and second colors;
An extraction step of sequentially extracting each pixel included in a predetermined processing region of the color image;
A defining step including a conversion source point on the color coordinate corresponding to the color of the pixel extracted in the extraction step and defining a conversion line along the reference line on the color coordinate;
A conversion step of calculating a post-conversion point obtained by moving the conversion source point by a predetermined movement amount according to the conversion line defined in the defining step;
A replacement step of replacing the color of the extracted pixel with a color corresponding to the post-conversion point calculated in the conversion step;
Including repeatedly executing the defining step, the converting step, and the replacing step until all pixels included in the processing region are extracted in the extracting step ,
The reference line is defined as a linear expression on the color coordinates,
The program in which the conversion line is defined to be parallel to the reference line on the color coordinate . An image processing apparatus for reducing a color difference between pixels with respect to a color image composed of a plurality of pixels,
First defining means for defining, on the color coordinates, a reference line including two reference points on the color coordinates corresponding to the selected first and second colors;
Extraction means for sequentially extracting each pixel included in a predetermined processing region of the color image;
Second defining means for defining on the color coordinates a conversion line that includes a conversion source point on the color coordinates corresponding to the color of the pixel extracted by the extracting means, and that is along the reference line;
Conversion means for calculating a post-conversion point obtained by moving the conversion source point by a predetermined amount of movement according to the conversion line defined by the second defining means;
E Bei and replacement means for replacing the color of pixels the extracted color corresponding to the converted point calculated by said converting means,
The reference line is defined as a linear expression on the color coordinates,
The image processing apparatus, wherein the conversion line is defined to be parallel to the reference line on the color coordinates .