JP3557115B2 - Image filter determination method and apparatus, and recording medium recording program for executing the processing - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to an image processing technique, and more particularly, to a technique for processing a pixel group having a predetermined color characteristic.
[0002]
[Prior art]
In image processing, there are cases where it is desired to perform processing only on pixels having color characteristics close to a specific color among a plurality of pixels included in an original image. In such a case, processing is often performed using different processing parameters according to the color of each pixel.
[0003]
As a method of using different processing parameters according to the color of each pixel, for example, there is a method disclosed in Japanese Patent Application Laid-Open No. 5-300375 published by the present applicant. In this technique, the values of the R (red) component, G (green) component, and B (blue) component of each pixel are temporarily converted into values of hue, saturation, and lightness suitable for human perception. Pixels having color characteristics close to a specific color are specified. Further, image processing such as sharpness processing is performed by setting processing parameters based on the specified pixels.
[0004]
[Problems to be solved by the invention]
However, in the prior art, as described above, it is time-consuming to determine the processing parameters for image processing after the R, G, and B color components are once converted into values such as hues. There was a problem.
[0005]
SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems in the conventional technology, and has as its object to provide a technology capable of easily determining a processing parameter suitable for predetermined image processing.
[0006]
[Means for Solving the Problems and Their Functions and Effects]
In order to solve at least a part of the problems described above, a method of the present invention is a method of determining image processing parameters for performing predetermined image processing on an original image,
(A) selecting a pixel group having a predetermined color characteristic as a processing target pixel group by using color components of R, G, and B components of each pixel included in the original image;
(B) determining the image processing parameter for performing the predetermined image processing for each pixel according to a predetermined index value based on the color component of each pixel included in the processing target pixel group;
It is characterized by having.
[0007]
In this method, a processing target pixel group having predetermined color characteristics is selected using three color components of R, G, and B. Further, the image processing parameters are determined for each pixel using the color component of each pixel included in the processing target pixel group. This makes it possible to easily determine a processing parameter suitable for the predetermined image processing.
[0008]
In the above method,
The processing target pixel group is a pixel group having flesh color characteristics,
The step (a) includes, among the R component, the G component, and the B component of each pixel included in the original image, a pixel group in which the R component is larger than the G component and the R component is larger than the B component. Selected as the processing target pixel group,
The step (b) may include a step of determining the index value based on a difference between an R component and a B component of each pixel included in the processing target pixel group.
[0009]
This makes it possible to easily select a pixel group having flesh color characteristics. Further, an index value for each pixel included in the processing target pixel group can be easily obtained. Therefore, the image processing parameters can be easily determined.
[0010]
In addition, in this specification, "skin color" means a slightly reddish light yellow.
[0011]
In the above method,
It is preferable that the image processing parameter is determined so that it becomes a peak when the index value is substantially equal to a predetermined reference value, and changes monotonically as the difference between the index value and the predetermined reference value increases. .
[0012]
This makes it possible to determine an image processing parameter corresponding to the color component of each pixel for a pixel group having flesh color characteristics.
[0013]
In the above method,
The predetermined reference value is:
Preferably, the determination is made based on a difference between an R component and a B component of a pixel group including at least one pixel specified by a user in the original image.
[0014]
In this way, the image processing parameters can be determined based on the pixels having the desired “skin color”.
[0015]
The recording medium of the present invention is a computer-readable recording medium that records a computer program for determining image processing parameters in order to perform predetermined image processing on an original image,
A function of selecting a pixel group having predetermined color characteristics as a processing target pixel group using the R component, the G component, and the B component of each pixel included in the original image;
A function of determining the image processing parameter for performing the predetermined image processing for each pixel according to a predetermined index value based on the color component of each pixel included in the processing target pixel group,
Is recorded on a computer program for causing a computer to realize the above.
[0016]
Even when the computer program recorded on such a recording medium is executed by a computer system, the computer program has the same operation and effect as the above method, and easily determines processing parameters suitable for predetermined image processing. Becomes possible.
[0017]
Other aspects of the invention
The present invention also includes the following aspects. A first aspect is an image processing parameter determination device that determines image processing parameters for performing predetermined image processing on an original image,
A processing target pixel group selection unit that selects a pixel group having predetermined color characteristics as a processing target pixel group by using color components of an R component, a G component, and a B component of each pixel included in the original image;
An image processing parameter determining unit that determines the image processing parameter for performing the predetermined image processing for each pixel, according to a predetermined index value based on the color component of each pixel included in the processing target pixel group,
It is characterized by having.
[0018]
A second aspect is a program supply device that supplies, via a communication path, a computer program that causes a computer to execute each step or the function of each unit of the above invention.
[0019]
BEST MODE FOR CARRYING OUT THE INVENTION
Next, embodiments of the present invention will be described based on examples. FIG. 1 is a block diagram illustrating an example of an image processing apparatus to which the present invention has been applied. This device includes a CPU 100 and a bus line 100a. An image data memory 110, a ROM 120, a display unit 130, a magnetic disk 140, and an input device 150 including a mouse and a keyboard are connected to the bus line 100a. The RAM 190 is connected to the bus line 100a. In the RAM 190, a program for realizing the functions of the processing target pixel group selection unit 191, the parameter determination unit 192, the reference value setting unit 193, and the image processing unit 194 is stored. This image processing apparatus may be connected to a transmission path to form a general network system.
[0020]
A computer program that realizes the functions of the units 191 to 194 in the RAM 190 is provided in a form recorded on a computer-readable recording medium such as a flexible disk or a CD-ROM. The computer reads the computer program from the recording medium and transfers it to an internal storage device or an external storage device. Alternatively, a computer program may be supplied to a computer via a communication path. When implementing the functions of the computer program, the computer program stored in the internal storage device is executed by the microprocessor of the computer. Further, the computer may read a computer program recorded on a recording medium and directly execute the computer program.
[0021]
In this specification, the computer is a concept including a hardware device and an operation system, and means a hardware device that operates under the control of the operation system. In the case where an operation system is unnecessary and a hardware device is operated by an application program alone, the hardware device itself corresponds to a computer. The hardware device includes at least a microprocessor such as a CPU and means for reading a computer program recorded on a recording medium. The computer program includes a program code that causes such a computer to realize the functions of the above-described units. Some of the functions described above may be realized by an operation system instead of the application program.
[0022]
Note that the “recording medium” in the present invention includes a flexible disk, a CD-ROM, a magneto-optical disk, an IC card, a ROM cartridge, a punched card, a printed matter on which a code such as a barcode is printed, a computer internal storage device (RAM And a computer-readable medium such as an external storage device.
[0023]
FIG. 2 is a flowchart illustrating a procedure for determining processing parameters when performing image processing. In this embodiment, a case will be described in which a pixel group having a color characteristic close to skin color is set as a processing target pixel group and smoothing (smoothing) processing is performed. Therefore, the processing parameters obtained by the processing of FIG. 2 are processing parameters used when performing the smoothing processing of the skin color area. As described later, the effect of the smoothing process increases as the processing parameter increases, and the effect of the smoothing process decreases as the processing parameter decreases.
[0024]
In step S101, the i-th pixel data is selected as target pixel data from all the pixel data included in the original image data. This original image data is read from the magnetic disk 140 shown in FIG. 1 and stored in the image data memory 110. Each of the three color components R, G, and B of each pixel data is composed of 8 bits.
[0025]
In step S102, the relationship between the respective color components for the i-th pixel data is checked to determine whether or not the i-th pixel data has a predetermined color characteristic. That is, as shown in step S102 of FIG. 2, the R component of the i-th pixel data is larger than the G component (“Ri> Gi”), and the R component is larger than the B component (“Ri> Bi”). Check whether the condition is satisfied. A pixel that satisfies this condition is determined to be a pixel having color characteristics close to skin color.
[0026]
FIG. 3 is an explanatory diagram showing a color region satisfying the condition of step S102 in the RGB color system. The color system shown in FIG. 3 includes an R axis representing an R component, a G axis representing a G component, and a B axis representing a B component. Furthermore, the Y axis representing the Y (yellow) component is located in the RG plane, the C axis representing the C (cyan) component is located in the GB plane, and the M (magenta) component is located in the BR plane. Is shown in the figure. Also, in FIG. 3, a symbol “W” representing white is shown on a straight line formed by points passing through the coordinate origin O of the color system and having three color components of R, G, and B having the same value. I have. Note that the coordinate origin O corresponds to black (black). Since the condition of step S102 is an area where “R> G” and “R> B”, it corresponds to the shaded area shown in FIG. That is, the area inside the square pyramid SC hatched in FIG. 3 is an area showing a color characteristic close to flesh color.
[0027]
FIG. 4 is an explanatory diagram showing a color region satisfying the condition of step S102 in a hue circle. The hue circle shown in FIG. 4 shows the hues in the NCS (Natural Color System) color space. The hue circle includes an R axis representing an R component, a G axis representing a G component, a B axis representing a B component, and a Y axis representing a Y component. The center point of the hue circle indicates an achromatic color, and there are a white point and a black point in two directions passing through the center and perpendicular to the paper surface. In the hue circle of FIG. 4, a region having a color characteristic close to the flesh color of “R> G” and “R> B” corresponds to a hatched region shown in FIG. 4.
[0028]
In step S102 of FIG. 2, if the i-th pixel data does not satisfy the conditions of “Ri> Gi” and “Ri> Bi”, in other words, if the i-th pixel data does not have color characteristics close to skin color, Proceeds to step S103. At this time, in step S103, the processing parameter pi for the i-th pixel data is determined to be “pi = 0”.
[0029]
In step S102 of FIG. 2, if the i-th pixel data satisfies the condition of “Ri> Gi” and “Ri> Bi”, in other words, if the i-th pixel data is pixel data having a color characteristic close to skin color, Proceed to S104.
[0030]
Note that the processing of steps S101 to S102 is executed by the processing target pixel group selection unit 191 shown in FIG.
[0031]
In step S104, an index value Vi for the i-th pixel data is calculated. The index value Vi for the i-th pixel data is a value for determining the processing parameter pi for the i-th pixel data. The index value Vi of the present embodiment is determined by “Vi = Ri−Bi” as shown in step S104 of FIG. In this embodiment, since each color component of the pixel data is represented by 8 bits as described above, it can take a value from “0” to “255”. Further, as can be seen from step S102, the i-th pixel data satisfies the relationship of “Ri> Bi”. Therefore, the index value Vi of the i-th pixel data is a value of “1” or more and “255” or less.
[0032]
After calculating the index value Vi, in steps S105 to S109, the processing parameter pi according to the index value Vi is determined. Note that the above-described processing in step S103 and steps S104 to S109 are executed by the parameter determination unit 192 in FIG.
[0033]
FIG. 5 is an explanatory diagram showing the relationship between the index value V calculated in step S104 of FIG. 2 and the processing parameter p determined in steps S105 to S109. As shown in FIG. 5, the processing parameter p becomes “1” when the index value V is equal to the reference value W, and changes linearly to “0” as the index value V moves away from the reference value W. This reference value W is a value serving as a reference for “skin color”. Therefore, the index value V indicates that the closer to the reference value W, the closer to the flesh color. The reference value W is set by the user using the input device 150 (FIG. 1) in the reference value setting unit 193.
[0034]
In this embodiment, each color component of the pixel data is represented by 8 bits, and the target value is an original image in which a yellow race such as Japanese is expressed. Is set to The area where the index value V is close to the reference value W (about “80”) corresponds to the “skin color” area FCA shown in the hue circle of FIG. The reference value W may be changed according to the skin color of the person represented in the target original image.
[0035]
If image processing is performed using the processing parameters that change continuously as shown in FIG. 5, substantially continuous even at the boundary between a portion having color characteristics close to skin color and a portion not having color characteristics close to skin color. Processing can be performed.
[0036]
In step S105 of FIG. 2, first, it is determined whether or not the relationship between the index value Vi and the reference value W for the i-th pixel data satisfies the condition of “Vi ≦ W”. If the index value Vi satisfies the condition of “Vi ≦ W” in step S105, the process proceeds to step S106. At this time, in step S106, the processing parameter pi for the i-th pixel data is determined to be “pi = Vi / W”. The relationship between the index value Vi and the processing parameter pi at this time corresponds to a linear region having a positive slope of “0 <V ≦ W” in FIG. On the other hand, if the index value Vi does not satisfy the condition of “Vi ≦ W” in step S105, the process proceeds to step S107.
[0037]
In step S107, it is determined whether or not the relationship between the index value Vi and the reference value W for the i-th pixel data satisfies the condition of “W <Vi ≦ 2 · W”. If the index value Vi satisfies the condition of “W <Vi ≦ 2 · W” in step S107, the process proceeds to step S108. At this time, in step S108, the processing parameter pi for the i-th pixel data is determined to be “pi = 2-Vi / W”. Note that the relationship between the index value Vi and the processing parameter pi at this time corresponds to a linear area having a negative slope of “W <V ≦ 2 · W” in FIG. On the other hand, if the index value Vi does not satisfy the condition of “W <Vi ≦ 2 · W” in step S108, the process proceeds to step S109. At this time, the index value pi for the i-th pixel data is determined to be “pi = 0”. Note that the relationship between the index value Vi and the processing parameter pi at this time corresponds to the area of “2 · W <V” in FIG.
[0038]
When the processing parameter pi for the i-th pixel data is determined as described above, in step 110, “i” is incremented by one. At this time, the process returns to step S101, and the (i + 1) -th pixel data as the next pixel data is selected as the pixel data of interest, and the processing of steps S102 to S110 is performed. As described above, by executing the processing of steps S101 to S110 with all pixel data included in the original image data as target pixel data, a processing parameter p corresponding to each pixel data can be obtained.
[0039]
In FIG. 5, in a region where the processing parameter p is substantially “0”, in a region where the index value V is close to “2 · W”, the R component is sufficiently larger than the B component. As the pixel data of this area, for example, relatively bright red pixel data corresponds. On the other hand, in a region where the index value V is close to “0”, the R component and the B component have substantially equal values. As the pixel data in this area, for example, pixel data having a gray color corresponds. Gray is a color when the R component, the G component, and the B component are substantially equal. As described above, the processing parameter p approaches “0” for pixel data that does not have a color that is very close to the skin color, so that the effect of the smoothing processing can be reduced.
[0040]
By the way, pixel data having a magenta color also corresponds to pixel data in an area where the index value V is close to “0”. Note that magenta is a color when the R component and the B component are substantially equal and the G component is small. Therefore, the processing parameter p becomes a value close to “0” for the pixel of the lip or lipstick part close to the color of magenta (red), so that the effect of the smoothing processing can be reduced. Therefore, when a human face is represented in the original image, it is possible to obtain a good processed image in which the skin portion is smooth and the lip (lipstick) portion is sharp by the smoothing process. It becomes.
[0041]
Also, in the original image, when the pixel region of “R> G” and the pixel region of “R <G” are adjacent to each other, the index value V may suddenly change at the boundary between the two pixel regions. is there. At this time, the value of the processing parameter p depending on the index value V also changes suddenly. For example, the processing parameter p may be “1” in a pixel area where “R> G”, and the processing parameter p may be “0” in a pixel area where “R <G”. In such a case, the smoothing process is discontinuous. However, the color of the pixel data in which the R component and the G component are substantially equal to each other corresponds to a yellow area near the Y axis as shown by the hue circle in FIG. Generally, a change in image quality near the Y-axis of the color wheel, particularly a change in sharpness, is not so noticeable. Therefore, even at the boundary between the pixel region of “R> G” and the pixel region of “R <G”, the processed image appears to have been successfully smoothed.
[0042]
In the above description, a linearly changing relationship as shown in FIG. 5 is used as the relationship between the index value V and the processing parameter p suitable for the smoothing process of the skin color portion, but other relationships are used. Is also good. FIG. 6 is an explanatory diagram illustrating another relationship between the index value V and the processing parameter p. In FIG. 6, the processing parameter p becomes “1” when the index value V is equal to the reference value W, and changes to “0” in a substantially Gaussian distribution as the index value V moves away from the reference value W. Even using such a relationship, it is possible to obtain a processing parameter p suitable for the smoothing processing of the skin color portion. That is, the processing parameter p may be a parameter that peaks when the index value V is equal to the reference value W, and changes monotonically as the difference between the index value V and the reference value W increases.
[0043]
In the above description, the case where the reference value W is set in advance has been described, but the reference value W may be changed according to the original image. In this case, for example, among all the pixels included in the original image, a pixel group including at least one pixel having a color to be set as a reference skin color is designated as a reference pixel group, and a reference value W is calculated. do it. At this time, the average value of the difference (R−B) between the R component and the B component of each pixel included in the reference pixel group is obtained as the reference value W. The designation of the reference pixel group may be performed before starting the processing in FIG. The designation of the reference pixel group can be performed on the original image displayed on the screen of the display unit 130 (FIG. 1) using the input device 150 such as a mouse. Alternatively, the process of step S102 shown in FIG. 2 is performed in advance on all the pixels included in the original image, and after selecting a pixel group having a color characteristic close to the skin color, a reference pixel group is selected from the selected pixel group. It may be specified. This has the advantage that the reference pixel group can be easily specified.
[0044]
When the processing parameter p is determined for each pixel of the original image as described above, a smoothing process is executed by the image processing unit 194 shown in FIG. The smoothing process according to the present embodiment is executed sequentially with all pixels included in the original image as target pixels. Alternatively, in step S102 in FIG. 2, the smoothing process may be performed only on the pixels determined to have the color characteristics close to the skin color. In this case, the processing in step S103 can be omitted.
[0045]
In the smoothing processing of the present embodiment, the smoothing filter used for the smoothing processing is changed according to the processing parameter p.
[0046]
FIG. 7 is an explanatory diagram showing the relationship between the processing parameter p and the smoothing filter. FIGS. 7A-1 and 7A-2 show cases where the filter size of the smoothing filter changes according to the processing parameter p. That is, when the processing parameter p is small, the filter size is small, and when the processing parameter p is large, the filter size is large. Note that the smoothing filters shown in FIGS. 7A-1 and 7A-2 are, for example, average filters in which all filter coefficients are 1.
[0047]
The filter size ((S (p) × S (p)) pixels) of the smoothing filter shown in FIGS. 7A-1 and 7A-2 is obtained, for example, by the following equation (1).
[0048]
S (p) = 2 · m · p + 1 (1)
[0049]
Here, m is a positive constant indicating the degree of smoothing, and the greater the value of m, the greater the effect of the smoothing process.
[0050]
In FIGS. 7A-1 and 7A-2, "2" is set as the constant m in equation (1). Therefore, when the processing parameter p is “0” (FIG. 7A-1), the filter size is (1 × 1) pixels, which is equivalent to substantially not performing the smoothing processing. When the processing parameter p is “1” (FIG. 7A-2), the filter size is (9 × 9) pixels.
[0051]
If smoothing processing is performed on pixels having color characteristics close to skin color using such a smoothing filter, the effect of the smoothing processing can be enhanced for skin color portions where the processing parameter p is large. In addition, the effect of the smoothing process can be reduced for parts such as gray and red where the processing parameter p is small.
[0052]
On the other hand, FIGS. 7 (B-1) and 7 (B-2) show the case where the filter size of the smoothing filter is constant, but the filter coefficient at the center of the smoothing filter changes according to the processing parameter p. I have. That is, when the processing parameter p is small, the central filter coefficient is large, and when the processing parameter p is large, the central filter coefficient is small.
[0053]
The filter coefficient CC (p) at the center of the smoothing filter shown in FIGS. 7 (B-1) and (B-2) is obtained, for example, by the following equation (2).
[0054]
CC (p) = − (n−1) · p + n (2)
[0055]
Here, n is a positive constant indicating the degree of smoothing. As n increases, the “weight” of the target pixel increases, in other words, the effect of the smoothing process decreases.
[0056]
7 (B-1) and (B-2), "20" is set as the constant n in the equation (2). Therefore, when the processing parameter p is “0” (FIG. 7B-1), the center filter coefficient is “20”. Note that the smoothing process is also performed on the pixel whose processing parameter is “0”, but the effect of the smoothing process is small because the center filter coefficient is large. When the processing parameter p is “1” (FIG. 7B-2), the center filter coefficient is “1”.
[0057]
Even when such a smoothing filter is used, the effect of the smoothing process is increased for a skin color portion where the processing parameter p is large, and the effect of the smoothing process is small for a gray or red portion where the processing parameter p is small. It is possible to do.
[0058]
As described above, in the present invention, it is determined whether each pixel has a predetermined color characteristic by using three color components of R, G, and B of each pixel data included in the original image data. I do. Further, for each pixel data having a predetermined color characteristic, a processing parameter is determined by obtaining an index value V based on the color component. This makes it possible to easily determine a processing parameter suitable for predetermined image processing.
[0059]
It should be noted that the present invention is not limited to the above-described examples and embodiments, and can be implemented in various modes without departing from the gist of the invention, and for example, the following modifications are possible.
[0060]
(1) In the above embodiment, as shown in FIG. 2, the index value Vi for the i-th pixel data is calculated by “Vi = Ri−Bi”, but the index value Vi is calculated using another equation. You may ask for it. For example, it may be obtained by “Vi = s · Ri / Bi”. Here, s is a positive constant. Even using this equation, it is possible to obtain the processing parameter p according to the relationship with the reference value W.
[0061]
(2) In the above embodiment, the case where the pixel group having the predetermined color characteristic is a pixel group having a color characteristic close to skin color has been described, but a pixel group having another color characteristic may be targeted.
[0062]
(3) In the above embodiment, the case where the smoothing process is performed using the determined processing parameter p has been described. However, a sharpness process, a color conversion process, or the like may be performed.
[0063]
(4) In the above embodiment, a part of the configuration realized by software may be replaced with hardware.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating an example of an image processing apparatus to which the present invention has been applied.
FIG. 2 is a flowchart illustrating a procedure for determining processing parameters when performing image processing.
FIG. 3 is an explanatory diagram showing a color region satisfying a condition of step S102 in an RGB color system.
FIG. 4 is an explanatory diagram showing a color region that satisfies a condition of step S102 in a hue circle.
FIG. 5 is an explanatory diagram showing a relationship between an index value V calculated in step S104 of FIG. 2 and a processing parameter p determined in steps S105 to S109.
FIG. 6 is an explanatory diagram showing another relationship between an index value V and a processing parameter p.
FIG. 7 is an explanatory diagram showing a relationship between a processing parameter p and a smoothing filter.
[Explanation of symbols]
100 ... CPU
100a ... bus line
110 ... Image data memory
120 ... ROM
130 Display unit
140 ... magnetic disk
150 ... input device
190 ... RAM
191... Target pixel group selection unit
192: Parameter determination unit
193: Reference value setting section
194 ... Image processing unit

Claims (9)

A method for determining an image filter for performing predetermined image processing on an original image,
(A) selecting a pixel group having a predetermined color characteristic as a processing target pixel group by using color components of R, G, and B components of each pixel included in the original image;
And (b) have groups Dzu on the color components of each pixel included in the target pixel group, for each pixel, a step of determining an index value that indicates that close to a predetermined reference color closer to a predetermined reference value,
(C) Using the index value for each pixel, obtain a parameter for determining the content of the image filter for each pixel, and use the parameter for each pixel to determine the content of the image filter for each pixel. The step of determining;
A method for determining an image filter , comprising: A method for determining an image filter according to claim 1 ,
Previous Stories step (a), of the R component and G component and the B component pixels included in the original image has, the R component is larger than the G component and a pixel group R component is larger than the B component the comprises the step of selecting as the processing target pixel group,
The method of determining an image filter , wherein the step (b) includes a step of determining the index value based on a difference between an R component and a B component of each pixel included in the processing target pixel group. A method for determining an image filter according to claim 2,
Before Kipa parameters, the index value becomes a peak when approximately equal to the predetermined reference value, the difference between the index value and the predetermined reference value is determined so as to change monotonously as the larger image How to determine the filter . The method for determining an image filter according to claim 3, wherein
The predetermined reference value, before one of Kihara image is determined based on the difference between the R component and the B component of a pixel group including at least one pixel is designated by the user, the method for determining the image filter. A computer-readable recording medium that records a computer program that causes a computer to determine an image filter for performing predetermined image processing on an original image,
A function of selecting a pixel group having predetermined color characteristics as a processing target pixel group using the R component, the G component, and the B component of each pixel included in the original image;
And have groups Dzu on the color components of each pixel included in the target pixel group, for each pixel, a function for obtaining an index value that indicates that close to a predetermined reference color closer to a predetermined reference value,
A function for determining a parameter for determining the content of the image filter for each pixel using the index value for each pixel, and determining the content of the image filter for each pixel using the parameter for each pixel When,
And a computer-readable recording medium on which a computer program for causing a computer to realize the above. An image filter determination device that determines an image filter for performing predetermined image processing on an original image,
  A processing target pixel group selection unit that selects a pixel group having predetermined color characteristics as a processing target pixel group by using color components of an R component, a G component, and a B component of each pixel included in the original image;
  An index value determination unit that calculates an index value indicating that the closer to a predetermined reference value is closer to a predetermined reference color, for each pixel, based on the color component of each pixel included in the processing target pixel group,
  Using the index value for each pixel, obtain a parameter for determining the content of the image filter for each pixel, and use the parameter for each pixel to determine the content of the image filter for each pixel. A filter determining unit;
An image filter determining apparatus, comprising: The image filter determination device according to claim 6,
  The processing target pixel group selection unit is configured to, among the R component, the G component, and the B component included in each pixel included in the original image, select a pixel whose R component is larger than the G component and whose R component is larger than the B component Selecting a group as the pixel group to be processed;
  The image filter determination device, wherein the index value determination unit determines the index value based on a difference between an R component and a B component of each pixel included in the processing target pixel group. The image filter determination device according to claim 7,
  The parameter is determined such that the parameter peaks when the index value is substantially equal to the predetermined reference value, and changes monotonously as the difference between the index value and the predetermined reference value increases. apparatus. The image filter determining apparatus according to claim 8, wherein
  The image filter determination device, wherein the predetermined reference value is determined based on a difference between an R component and a B component of a pixel group including at least one pixel specified by a user in the original image.

Images