JP2004120382A - Apparatus, method and program for subtractive color processing - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an apparatus, a method and a program for subtractive color processing which realize preventing face color of a person contained in a picture being a target of subtractive color processing from being changed. <P>SOLUTION: The apparatus for subtractive color processing comprises a face extraction processing section 61 extracting a region showing the face image of the person from a target image of the subtractive color processing; a face color average value calculation processing section 62 calculating the average value of colors contained in the face image region; a weighting processing section 63 setting the weighting factor of each color having a value close to the face color average value calculated by the face color average value calculation processing section 62; a face color information storage section 51 storing information on each color having the weighting factor set by the weighting processing section 63; a color palette preparation section 52 preparing a color palette containing representative colors determined based on a result applied with statistical processing by using face color information stored in the face color information storage section 51 with respect to the color information on the image being the target of the subtractive color processing; and a color replacement section 53 replacing the color of each pixel of the picture being the target of the substrative color processing with any color contained in the color palette. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a color reduction processing apparatus, a color reduction processing method, and a color reduction processing program for converting each pixel color of a color reduction processing target image into an image having a smaller number of colors than an original image by replacing the color of each pixel with a color included in a predetermined color palette. About.
[0002]
[Prior art]
In recent years, images of digital data taken with a digital camera or the like are transferred to an image distribution server using a mobile phone or a personal computer, or an image distribution server in which an image distributor prepares various image data for distribution has been operated. An image distribution system that distributes images stored in these image distribution servers to mobile terminals such as mobile phones and personal digital assistants (PDAs) has been put into practical use.
[0003]
In this type of image distribution system, there are strong restrictions on the memory capacity and display capability of the mobile terminal, and the memory capacity and display capability differ depending on the type of mobile terminal, so when distributing images from the image distribution server to the mobile terminal According to the memory capacity and display capability of the distribution-destination portable terminal, processing is performed to reduce the data amount by reducing the number of images and the number of colors. For example, after performing image reduction processing (thinning processing) that reduces the number of pixels to a predetermined number of pixels such as 130 × 130 pixels and 90 × 90 pixels, the number of colors included in the image is further increased to 256 colors, 128 colors, 16 Color reduction processing for reducing the number of colors such as colors to a predetermined number is performed.
[0004]
Conventionally, as this kind of color reduction processing method, a fixed type limited color display method (fixed palette color reduction method) and an adaptive type limited color display method (free palette color reduction method) are known.
[0005]
The fixed limited color display method uses one color palette consisting of a predetermined number of representative colors that are uniformly selected from colors that can be reproduced by a display device, and uses the color palette for the color of each pixel of the color reduction target image. This is a color reduction processing method that replaces the closest color included. This method does not need to create a color palette for each color reduction target image, so it can perform color reduction processing at high speed, but there is a bias in the colors in the image, such as a face image or autumnal image. In some cases, only some of the colors included in the color palette are used, and the remaining colors are not used at all. For example, only 50 colors may be used even though a color palette composed of 256 representative colors is used. In such a case, the color image representation after the color reduction processing is of low quality.
[0006]
The adaptive limited color display method creates a color palette composed of a predetermined number of representative colors suitable for expressing the color reduction processing target image based on the color information of the color reduction processing target image. This is a color reduction processing method in which the color of a pixel is replaced with the closest color included in the color palette. This method needs to create a color palette for each color reduction target image, so the processing speed is slower than the fixed limited color display method, but it uses a color palette suitable for each color reduction target image. High-quality color image representation is possible. As a technique using the adaptive limited color display method, there is a technique of creating a color palette based on statistical information regarding the color of each pixel constituting a color reduction processing target image when performing color reduction processing (Patent Document 1). reference.).
[0007]
[Patent Document 1]
JP 2001-309190 A
[0008]
[Problems to be solved by the invention]
In the adaptive limited color display method described above, the appearance frequency of each color in the color reduction processing target image is obtained by statistical processing using a histogram or the like, and a predetermined number of colors are given priority for each color reduction processing target image. A color palette is created by selecting representative colors.
[0009]
For this reason, for example, in the case of an image in which a person appears small against a background of autumn leaves or a green field, the person's face is the main subject, but the color contained in the person's face Since the appearance frequency of the colors included in the background is overwhelmingly high, a color palette including more colors included in the background than the colors included in the person's face is created. When color reduction processing is performed using this color palette, the image of the background part is expressed smoothly in many colors, so the degree of image quality degradation is small, but the background color is important for human faces. Since it is expressed by a small number of tasteful colors, the degree of degradation of image quality increases. In particular, when a large color reduction process is performed, important colors such as the skin color of a person's face may be deleted from the color palette, and the color may change significantly with respect to the original image. For example, in the case of an image in which a person is shown small with autumn leaves in the background, the face color of the person becomes a reddish color as a result of the color reduction processing.
[0010]
The present invention has been made in view of the above circumstances, and an object of the present invention is to prevent the color of a person's face, which is an important color in a color reduction target image, from being deleted from the color palette when performing color reduction processing. Another object of the present invention is to provide a color reduction processing apparatus, a color reduction processing method, and a color reduction processing program capable of performing color reduction processing while suppressing deterioration in image quality as much as possible.
[0011]
[Means for Solving the Problems]
In order to achieve the above object, a color reduction processing apparatus according to the present invention includes a color palette creation means for creating a color palette having a predetermined number of representative colors, and the color of each pixel of the color reduction target image using the color palette. A color replacement processing unit that replaces any one of the colors included in the color palette, and a facial color that extracts a facial image area of the subject from the color-reduction processing target image and generates facial color information related to the colors included in the facial image area Information generating means, wherein the color palette creating means creates a color palette including representative colors determined based on the face color information.
[0012]
In the color reduction processing apparatus of the present invention configured as described above, first, a human face image area is extracted from a color reduction processing target image, and facial color information relating to colors included in the face image area is generated. Next, a color palette composed of a predetermined number of representative colors including the representative color determined based on the face color information is created, and the color of each pixel of the color reduction target image is changed to a color included in the created color palette. Replacement color replacement processing is performed. Therefore, since the representative color of the color palette is determined based on the facial color information when performing the color reduction processing, it is possible to prevent the human facial color, which is an important color of the color reduction target image, from being easily deleted from the color palette. In addition, it is possible to avoid a significant change in the face color of a person due to color reduction, and it is possible to perform a color reduction process that suppresses deterioration in image quality as much as possible.
[0013]
Further, in the color reduction processing apparatus according to the present invention, the face color information generation unit includes a face color average value calculation unit that calculates an average value of colors included in the face image region, and each color near the face color average value. Weighting processing means for setting weighting coefficients, respectively, and the color palette creating means performs statistical processing on the color information of the color reduction target image using the weighting coefficients set by the weighting processing means. A color palette including a representative color determined based on the result is created.
[0014]
In the color reduction processing apparatus of the present invention configured as described above, first, a face image area of a person is extracted from a color reduction processing target image, a facial color average value is calculated, and a weighting coefficient is set for each color in the vicinity of the facial color average value. . Next, a color palette including representative colors determined based on the result of performing statistical processing on the color information of the color reduction processing target image using the weighting coefficient set by the weighting processing unit is created. Therefore, when performing the color reduction process, by setting a weighting factor for the color near the face color average value, the face color of the person, which is an important color of the color reduction target image, can be easily extracted from the color palette based on the appearance frequency. It can be prevented from being deleted, it can be avoided that the face color of the person changes greatly due to color reduction, and it is possible to perform a color reduction process that suppresses deterioration of image quality as much as possible.
[0015]
In the color reduction processing device according to the present invention, the weighting processing unit sets a weighting factor for each color in the vicinity of the face color average value according to a ratio of the face image area to the entire color reduction processing target image. It is characterized by doing. Therefore, when the area occupied by the face image area is small, colors other than the face color are easily set as the representative colors of the color palette.
[0016]
In the color reduction processing apparatus according to the present invention, the weighting processing unit sets a weighting factor for each color in the vicinity of the face color average value according to a distance from the face color average value. Therefore, since the weighting coefficient set for the color close to the face color average value is high, the color close to the face color average value is easily set as the representative color of the color palette.
[0017]
Further, the color reduction processing method according to the present invention extracts a face image area of a subject from a color reduction processing target image, generates a face color information related to a color included in the face image area, and includes the face color information in the face color information. A color palette creation step for creating a color palette including representative colors determined based on the color palette, and a color palette created in the color palette creation step, and using the color palette, the color of each pixel of the color reduction target image is stored in the color palette. And a color replacement processing step for replacing with any of the included colors.
[0018]
Therefore, since the representative color of the color palette is determined based on the facial color information when performing the color reduction processing, it is possible to prevent the human facial color, which is an important color of the color reduction target image, from being easily deleted from the color palette. In addition, it is possible to avoid a significant change in the face color of a person due to color reduction, and it is possible to perform a color reduction process that suppresses deterioration in image quality as much as possible.
[0019]
Further, in the color reduction processing method according to the present invention, the face color information generating step includes a face color average value calculating step for calculating an average value of colors included in the face image area, and each color near the face color average value. Weighting processing steps for setting weighting factors, respectively, and the color palette creating step performs statistical processing on the color information of the color reduction target image using the weighting factors set in the weighting processing step. The appearance frequency calculating step for obtaining the appearance frequency of each color in the color reduction processing target image, and the representative color for selecting the color having the higher appearance frequency as the representative color of the color palette based on the result obtained in the appearance frequency calculating step. And a selection step.
[0020]
Therefore, when performing the color reduction process, by setting a weighting factor for the color near the face color average value, the face color of the person, which is an important color of the color reduction target image, can be easily extracted from the color palette based on the appearance frequency. It can be prevented from being deleted, it can be avoided that the face color of the person changes greatly due to color reduction, and it is possible to perform a color reduction process that suppresses deterioration of image quality as much as possible.
[0021]
Furthermore, the color reduction processing program according to the present invention is for realizing the function of each means in the color reduction processing device according to any one of claims 1 to 4 by a computer.
[0022]
Therefore, since the representative color of the color palette is determined based on the facial color information when performing the color reduction processing, it is possible to prevent the human facial color, which is an important color of the color reduction target image, from being easily deleted from the color palette. In addition, it is possible to avoid a significant change in the face color of the person due to the color reduction, and it is possible to realize a color reduction process that suppresses the deterioration of the image quality as much as possible by the computer.
[0023]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the drawings. In the description of the present embodiment, the color reduction processing apparatus and the color reduction processing method according to the present invention will be described in detail. However, the color reduction processing program according to the present invention realizes the functions of the respective units included in the color reduction processing apparatus. Therefore, the description about the program is included in the following description.
FIG. 1 is a block diagram showing a configuration example of an image distribution system using a color reduction processing apparatus according to an embodiment of the present invention.
[0024]
An image distribution system 1 according to this embodiment includes a personal computer (distribution source terminal device, hereinafter referred to as a PC) 10, a mobile phone (distribution destination terminal device) 20 as a mobile terminal, and an image distribution server (image conversion distribution). Device) 30 is interconnected via a communication network such as the Internet, and an image of digital data such as an image taken by the digital camera 2 is transferred to the image distribution server 30 using the PC 10, and the image distribution server 30 The image stored in the mobile phone 20 is distributed to the mobile phone 20.
[0025]
An application program (hereinafter referred to as an image distribution application) for transferring an image distributed to the mobile phone 20 to the image distribution server 30 is installed in the PC 10. A user of the PC 10 such as an image distributor can distribute an image from a large number of images such as an image taken from the digital camera 2 to the PC 10 or an image created by using image processing software by using an image distribution application. Select the desired image. When an image to be distributed is selected using the image distribution application and a transfer instruction is given, the PC 10 transfers the selected image to the image distribution server 30.
[0026]
The mobile phone 20 includes a communication unit 21 for communicating with other mobile phones and servers on the Internet via a base station (not shown), a display unit 22 for displaying characters and images, and various operators such as dial buttons. And a terminal information storage unit 24 that stores information such as the communication speed of the mobile phone 20 and the display capability (the number of pixels and the number of colors that can be expressed) of the display unit 22. . When the mobile phone 20 communicates with another mobile phone or a server on the Internet, the mobile phone 20 reads the terminal information from the terminal information storage unit 24 and notifies the communication destination.
[0027]
The image distribution server 30 performs a face extraction process on the pre-distribution image storage unit 31 that stores the image (original image) received from the PC 10 and the image read from the pre-distribution image storage unit 31 to calculate a facial color average value. Thereafter, a face color information processing unit 35 that weights colors included in the face image, a terminal information storage unit 32 that stores terminal information received from the mobile phone 20, and an image read from the pre-delivery image storage unit 31 A distribution image generation unit 33 that generates a distribution image by performing a compression process after performing pixel number reduction processing and color reduction processing based on terminal information read from the information storage unit 32, and generated by the distribution image generation unit 33 And a transfer unit 34 that transfers the delivered image to the mobile phone 20.
[0028]
Next, the face color information processing unit 35 will be described with reference to FIG. FIG. 2 is a block diagram illustrating a configuration example of the face color information processing unit in the present embodiment.
As shown in the figure, the face color information processing unit 35 includes a face extraction processing unit (face extraction unit) 61 that extracts a human face image region from an image read from the pre-delivery image storage unit 31, and a face extraction processing unit 61. Based on the calculation results of the face color average value calculation processing unit (face color average value calculation means) 62 for calculating the face color average value of the face image area extracted by the above, and the face color average value calculation processing unit 62, weight coefficients are set for each color. And a weighting processing unit (weighting processing means) 63. The face color average value calculation processing unit 62 and the weighting processing unit 63 constitute face color information generating means for generating face color information.
[0029]
The face extraction processing unit 61 extracts a human face image area from the image read from the pre-delivery image storage unit 31. As a method for extracting a face image area, for example, the technique disclosed in Japanese Patent Laid-Open No. 9-101579 “Face Area Extraction Method and Copying Condition Determination Method” by the present applicant can be applied. This technology determines whether or not the hue of each pixel of the captured image is included in the skin color range and divides it into a skin color region and a non-skin color region, and detects edges in the image to detect each pixel in the image. Are classified into edge portions or non-edge portions. Then, an area surrounded by pixels that are located in the skin color area and classified as a non-edge part and surrounded by pixels that are determined to be edge parts is extracted as a face candidate area, and the extracted face candidate area becomes a human face. It is determined whether the region corresponds to the region, and the face image region is extracted based on the determination result.
[0030]
The face color average value calculation processing unit 62 expresses an average value when colors included in the face image area extracted by the face extraction processing unit 61 are expressed in a color space such as RGB or Lab (hereinafter referred to as a face color average value). Is calculated. It should be noted that although processing is simple when RGB is used as the color space used for calculating the face color average value, it is preferable to use Lab in order to increase the calculation accuracy.
[0031]
The weighting processing unit 63 sets weighting coefficients for colors near the face color average value, that is, colors that are close to the face color average value in the color space. For example, by setting the weighting coefficient of a certain color to 3, the appearance frequency of that color is counted as three times the actual appearance frequency. For this reason, even if the color has a low appearance frequency, the apparent appearance frequency can be increased by increasing the weighting factor, and the face color with a low appearance frequency is easily deleted from the color palette by statistical processing. Can be prevented. The weighting processing unit 63 increases the weighting coefficient for a color that is closer to the face color average value calculated by the face color average value calculation processing unit 62, that is, a color that is closer to the face color average value in the color space. Further, the weighting processing unit 63 decreases the weighting factor as the area ratio of the face image region to the entire image region is smaller, and increases the weighting factor as the area ratio is larger. As a result, when the area occupied by the face image area is small, colors other than the face color are easily set as the representative colors of the color palette. Furthermore, the weighting processing unit 63 notifies the face color information storage unit 51 in the color reduction processing unit 42 of the color near the face color average value and the weighting information that is information on the weighting coefficient set for each color as face color information.
[0032]
Next, the distribution image generation unit 33 will be described with reference to FIG. FIG. 3 is a block diagram illustrating a configuration example of the distribution image generation unit in the present embodiment.
As shown in the figure, the distribution image generation unit 33 includes a pixel number reduction processing unit 41 that reduces the number of pixels, a color reduction processing unit 42 that performs color reduction, a compression processing unit 43 that performs reversible or irreversible data compression, And a data capacity determination unit 44 that determines whether the data capacity of the output image is equal to or less than a predetermined value.
[0033]
The pixel number reduction processing unit 41 reduces the number of pixels of the image (original image) read from the pre-delivery image storage unit 31 to a predetermined number of pixels such as 130 × 130 pixels and 90 × 90 pixels (decimation processing). I do. At that time, the number of pixels after reduction (target value) is determined according to the number of pixels included in the terminal information read from the terminal information storage unit 33.
[0034]
The color reduction processing unit 42 is included in the image by replacing the color of each pixel of the image after being processed by the pixel number reduction processing unit 41 with the closest color among the colors included in the color palette created for each image. Color reduction processing is performed to reduce the number of colors to a predetermined number of colors such as 256 colors, 128 colors, and 16 colors.
[0035]
The compression processing unit 43 performs compression processing on the image (color-reduced image) processed by the color reduction processing unit 42 using a known compression method such as GIF or PNG. The data capacity determination unit 44 determines whether the data capacity of the image after data compression is equal to or less than the data capacity that can be processed in the communication path to the distribution destination and the portable terminal at the distribution destination and can be transmitted and displayed. The compressed image is transferred to the mobile phone 20 via the transfer unit 34 after the data volume is determined by the data volume determination unit 44.
[0036]
Since the transferable data capacity varies depending on the type of mobile phone and the communication carrier company, there is a possibility that transfer may be impossible or reception may be rejected due to the excess data capacity. If the data capacity determination unit 44 determines that the data capacity is exceeded, the image data is again subjected to further color reduction processing in the color reduction processing unit 42 and recompressed in the compression processing unit 43.
[0037]
As shown in FIG. 4, the color reduction processing unit 42 includes a facial color information storage unit 51 that stores facial color weighting information, a color palette creation unit (color palette creation unit) 52 that performs color palette creation processing, and a color palette. And a color replacement processing unit (color replacement processing means) 53 for replacing the color of each pixel in the original image.
[0038]
The face color information storage unit 51 stores the face color weighting information notified from the weighting processing unit 63 in the face color information processing unit 35. The color palette creation unit 52 is based on the result of performing statistical processing on the color information of the color reduction processing target image given from the pixel number reduction processing unit 41 using the facial color weighting information read from the facial color information storage unit 51. Create a color palette containing the representative colors determined in
[0039]
The color replacement processing unit 53 replaces the color of each pixel of the color reduction processing target image given from the pixel number reduction processing unit 41 with the closest color included in the color palette created by the color palette creation unit 52. Then, the color-reduction processing target image is converted into a color-reduced image composed of colors included in the color palette. The reduced color image is compressed by the compression processing unit 43 and then transferred to the mobile phone 20.
[0040]
FIG. 5 shows an operation procedure of the color palette creation unit 52. The color palette creating unit 52 first checks the color of each pixel constituting the color reduction processing target image given from the pixel number reduction processing unit 41 (step S1). Next, facial color weighting information is read from the facial color information storage unit 51 (step S3). Then, the color information of each pixel examined in step S1 is subjected to statistical processing using a histogram using the weighting information read in step S3, and the appearance frequency of each color in the color reduction processing target image is obtained (step S5). In this case, the appearance frequency is obtained by weighting the face color. A representative color having a higher appearance frequency, that is, a color in the vicinity of the face color average value is preferentially selected (step S7). In this way, a color palette (free palette) composed of representative colors preferentially including the important face color is created (step S9).
[0041]
As described above, a predetermined number of representatives including face color preferentially based on the result of performing statistical processing on the color information of the color reduction processing target image using the face color weighting information in the color palette creation unit 52. By creating a color palette composed of colors, it is possible to prevent the face color, which is an important color of the color reduction processing target image, from being easily deleted from the color palette based on the appearance frequency.
[0042]
Then, the color palette created by the color palette creating unit 52 is used, and the color replacement processing unit 53 replaces the color of each pixel of the color reduction processing target image with the color included in the color palette. The color reduction process can be performed without missing the face color of the person.
[0043]
Therefore, when the image to be distributed is an image in which a person is shown small with, for example, autumn leaves or green fields, it is an important color by weighting the color in the vicinity of the human face color average value. It is possible to prevent the face color from being easily deleted from the color palette so that the face color of the person is not lost as much as possible after the color reduction process. In the case of an image in which a person looks small with autumn leaves in the background, there is a problem that the face color of the person becomes a reddish color in the conventional color reduction processing method as a result of the color reduction processing. Such a problem can be solved by the color reduction processing method of the present invention described above.
[0044]
The function of the color reduction processing unit 42 can be realized by installing and operating a program for executing the color reduction processing by the color reduction processing method according to the present invention on a general-purpose server computer. That is, a step of creating a color palette composed of a predetermined number of representative colors preferentially including the face color, based on the result of statistical processing using the color information of the face color on the color information of the color reduction processing target image; A server computer having the function of the color reduction processing unit 42 can be realized by installing and executing a program including the step of replacing the color of each pixel of the color reduction processing target image with the color included in the color palette.
[0045]
As shown in the operation flow of FIG. 5, this program examines the color of each pixel constituting the color reduction target image in consideration of the weight color information, and performs the color reduction processing by statistical processing using a histogram or the like. It is desirable to have a step of obtaining the appearance frequency of each color in the target image and a step of preferentially selecting a color having a higher appearance frequency as the representative color.
[0046]
Note that if the human face is not included in the color reduction process target image, the face extraction process, the face color average value calculation process, and the weighting process are not performed, and normal statistical processing is performed to configure the color reduction process target image. The representative color of the color palette is selected by giving priority to the color having the higher appearance frequency of each pixel. Since the operation of the color reduction processing apparatus (color reduction processing method) when a human face is not included in the color reduction processing target image is the same as that of the conventional color reduction processing apparatus (color reduction processing method), description thereof is omitted.
[0047]
Further, when the color-reduced image after compression processing by the compression processing unit 43 needs to be further subjected to color-reduction processing by the color-reduction processing unit 42, the color of each pixel of the color-reduced image is added to the color palette created by the color palette creation unit 52. When replacing with the included color, the color that is closer to the face color average value calculated by the face color average value calculation processing unit 62, that is, the color that is closer to the face color average value in the color space, is replaced with a larger weight coefficient. It is desirable to do. By increasing the weighting coefficient for a color that is closer to the face color average value, it is possible to prevent a person's face color from being easily deleted by statistical processing using a histogram or the like.
[0048]
In the above embodiment, the example of the color reduction process using the color palette by the free palette color reduction method has been described. However, when the color reduction process is further performed after the color reduction once as described above, the color reduction process by the fixed palette color reduction method is used. Even in the color reduction processing using the color palette, when the number of colors in the color palette is reduced, the same effects can be obtained by applying the configuration and procedure of the above-described embodiment.
[0049]
As described above, according to the color reduction processing apparatus and the color reduction processing method in the present embodiment, a predetermined number is obtained based on the result of performing statistical processing on the color information of the color reduction target image using the facial color weighting information. A color palette consisting of the representative colors is created, and the color reduction processing to convert the color reduction processing target image into an image with fewer colors than the original image by replacing the color of each pixel of the color reduction processing target image with the color included in the color palette. When implementing, by setting a weighting factor for the color in the vicinity of the face color average value, it is possible to prevent the face color, which is an important color, from being easily deleted from the color palette based on the appearance frequency. It is possible to perform a color reduction process that suppresses deterioration as much as possible.
[0050]
【The invention's effect】
As described above, according to the present invention, it is possible to prevent the human face color, which is an important color of the color reduction target image, from being deleted from the color palette when performing the color reduction processing, and to suppress degradation of image quality as much as possible. Color reduction processing can be performed.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration example of an image distribution system using a color reduction processing apparatus according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating a configuration example of a face color information processing unit in the present embodiment.
FIG. 3 is a block diagram illustrating a configuration example of a distribution image generation unit in the present embodiment.
FIG. 4 is a block diagram illustrating a configuration example of a color reduction processing unit in the present embodiment.
FIG. 5 is a flowchart showing an operation procedure of a color palette creating unit in the present embodiment.
[Explanation of symbols]
1 Image distribution system
2 Digital camera
10 Personal computer (PC) (distribution source terminal device)
20 Mobile phone (distribution destination terminal device)
30 Image distribution server (Image conversion distribution device)
22 Display section
24 Terminal information storage unit
31 Pre-delivery image storage
32 Terminal information storage unit
33 Distribution image generator
34 Transfer section
35 Facial Color Information Processing Unit
41 Pixel reduction processing unit
42 Color Reduction Processing Unit (Color Reduction Processing Device)
43 Compression processing section
51 Facial color information storage unit
52 Color palette creation part
53 color replacement processor
61 Face extraction processing unit
62 Face color average value calculation processing unit
63 Weighting processing part

Claims (7)

Color palette creating means for creating a color palette having a predetermined number of representative colors;
Color replacement processing means for replacing the color of each pixel of the color reduction processing target image with any of the colors included in the color palette using the color palette;
Facial color information generating means for extracting a facial image area of a subject from the color reduction processing target image and generating facial color information relating to a color included in the facial image area;
The color pallet creating means creates a color pallet including representative colors determined based on the face color information. The face color information generating means includes a face color average value calculating means for calculating an average value of colors included in the face image area, a weighting processing means for setting a weighting factor for each color in the vicinity of the face color average value, Have
The color palette creating means includes a color palette including representative colors determined based on a result of performing statistical processing on the color information of the image to be subjected to color reduction processing using a weighting factor set by the weighting processing means. The color reduction processing apparatus according to claim 1, wherein the color reduction processing apparatus is created. The weighting processing unit sets a weighting factor for each color in the vicinity of the face color average value in accordance with a ratio of the face image region to the entire color reduction processing target image. Color reduction processing device. 4. The color reduction processing apparatus according to claim 2, wherein the weighting processing unit sets a weighting factor for each color in the vicinity of the face color average value according to a distance from the face color average value. A face color information generation step of extracting a face image area of the subject from the color reduction processing target image and generating face color information relating to a color included in the face image area;
A color palette creating step for creating a color palette including representative colors determined based on the face color information;
Using the color palette created in the color palette creating step, a color replacement processing step for replacing the color of each pixel of the color reduction processing target image with one of the colors included in the color palette;
A color-reduction processing method comprising: The face color information generating step includes a face color average value calculating step for calculating an average value of colors included in the face image area, and a weighting processing step for setting a weighting factor for each color in the vicinity of the face color average value, Including
The color palette creating step obtains the appearance frequency of each color in the color reduction target image by performing statistical processing on the color information of the color reduction target image using the weighting factor set in the weighting step. 6. An appearance frequency calculating step, and a representative color selecting step of selecting a color having a higher appearance frequency as a representative color of a color palette based on the result obtained in the appearance frequency calculating step. The color reduction processing method described. 5. A color reduction processing program for realizing the functions of each means in the color reduction processing apparatus according to claim 1 by a computer.

Images