JP2973710B2 - Limited color determination method and limited color determination device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for producing a natural image using a limited number of representative colors when displaying a digitized natural image on a display device with a small number of limited colors. The present invention relates to a limited color determination method and a limited color determination device for displaying.

[0002]

2. Description of the Related Art Generally, a color image used in a color display device is represented by RGB color components, and each color component is digitized by allocating an 8-bit memory. Therefore, in a full-color display for displaying the digitized color image, two pixels are required for each pixel.
4 bits of color, that is, simultaneously capable of displaying 2 ²⁴ colors, it is possible to display a natural image without discomfort.

On the other hand, CD-ROM, CD-I, CAD
(Computer Aided Design), CAI (Computer Assist
ed Instruction) and in the world of graphics,
^In many cases, a color display device capable of simultaneously displaying only a limited number of colors out of ²⁴ colors is used. In such a case, the capacity per pixel of the image memory is usually set to 8 bits or less. The number of possible colors is 2 ⁸
= 256. With a limited number of colors, such as 256 colors or less, it is difficult to display a natural image without deterioration, but there are many demands for making this possible. Also,
From the relationship of the data capacity of the display device and the like, the number of colors of the natural image is reduced so that the natural image does not deteriorate,
There are many requests to reduce the amount of image data.

For such a purpose, conventionally, for a color map display for selecting and displaying a small number of display colors from a large number of colors, a natural image is displayed in a limited number of colors while suppressing visual deterioration. Methods of displaying are known.

As the above method, for example, the colors R, G,
When represented by the values of the three primary colors of B, a space represented by a three-dimensional rectangular coordinate system having coordinates of the values of R, G, and B is defined as a color space. The color distribution of each pixel included is analyzed, and a large number of color regions in the color space where many colors included in the color image data are distributed, and a small number of display colors are selected from color regions with a small distribution. These display colors are registered in the color map, the display color closest to the color of each pixel of the color image data is assigned to each pixel as a representative color, and the color code registered as the representative color in the color map is stored in each image memory. In some cases, approximate display is performed by writing data at pixel positions. As one of the methods for selecting the representative colors as described above, Otsu's discriminant analysis (IEICE (D). J63-D, 4, pp. 349-356.
Showa 55-4. ).

The color space is not only a color space expressed by a combination of three primary colors of RGB, but also an L ^* u ^* v ^* color space and an L ^* a ^* b ^* color space in consideration of human color differences. CIE
(Recommended by the International Commission on Illumination) and the like, and the discriminant analysis method can be performed using these color spaces.

Further, a method utilizing the visual characteristics of humans and further applying the dither method (Journal of the Institute of Image Electronics Engineers of Japan, vol. 18, no. 5,
No. (1989) p. 293-301; Tajima, Ikeda), but this method does not consider the case of displaying limited colors with a small number of colors of 100 or less.

[0008]

Since the discriminant analysis method simply determines a display color based on a color distribution, an image created by determining a display color by the discriminant analysis method in the RGB color space is: For the human eye, natural images were not reproduced without any discomfort. For example, when a small number of representative colors are assigned to a portion where the color change of the image data is gradual, false contours may occur.In particular, since the human visual characteristics are sensitive to luminance, There has been a problem that when the change in the luminance gradation is represented by a small number of representative colors, false contours are noticeable.

Further, even if a discriminant analysis method is used in an L ^* u ^* v ^* color space or an L ^* a ^* b ^* color space in consideration of human color differences, a natural image is not always obtained when viewed from human eyes. It was not an approximation. Further, as a conventional method, as in the above-described discriminant analysis method, an image is created by selecting a limited number of display colors despite the fact that the state of the original image, for example, the color distribution is statistically different. There are two types of non-adaptive methods, and an adaptive method of selecting limited colors so as to be suitable for statistical properties such as the color distribution of the original image.

In the non-adaptive method, an image similar to the original image is not necessarily obtained when viewed from human eyes as in the case of the above-described discriminant analysis method, and there is a problem in image quality. In addition, in the adaptive method, the color distribution and the like of each image data are checked,
It is necessary to change the method of statistical processing according to the state of the color distribution, so that the processing is generally complicated and the processing time is long. The non-adaptive method that can maintain a certain level of image quality for general image data has been developed more than the adaptive method because of the advantage that the image data is not particularly selected and the processing time is short. Some natural images have very different statistical properties such as color distribution depending on the content, and the superiority of adaptive processing has not changed in terms of image quality.

The present invention has been made in view of the above-mentioned problems, and has been made in view of the fact that human vision is particularly sensitive to a change in luminance. It is an object of the present invention to provide a limited color determination method that can solve the problems between the non-adaptive method and the adaptive method.

[0012]

According to the present invention, a method for determining a limited color according to the present invention is arranged such that, in a uniform color space, a class number on a coordinate axis indicating a luminance component is different from a class number on a coordinate axis indicating a chromaticity component. set the number of class, the coordinates
A class that is three-dimensionally divided by the set class of axis
Color image represented by the coordinate system of the uniform color space for each data
Obtains a frequency distribution of data, divided the uniform color space by performing determine by analysis to the frequency distribution of each of said cluster
Each color area based on each color component of the divided color area.
The representative means for solving the above-mentioned problem is to obtain the representative color in the above and to approximate the color image data with the obtained representative color.

In the method for determining a limited color according to the present invention, when the frequency distribution is obtained in the uniform color space, the rank of the coordinate axis indicating the luminance component is set to be larger than the rank of the coordinate axis indicating the chromaticity component. The solution of the above-described problem is to select a representative color in which the luminance is enhanced.

Further, in the method for determining a limited color according to the present invention, when obtaining the frequency distribution in the uniform color space, the rank of the coordinate axis indicating the chromaticity component is set to be larger than the rank of the coordinate axis indicating the luminance component. The solution to the above-mentioned problem is to select a representative color in which chromaticity is emphasized. Also, limitations of the present invention
The color determination method is based on the image for each coordinate axis in the uniform color space.
The variance of the elementary distribution is obtained, and the ratio of the obtained variance between the coordinate axes.
The number of classes in the coordinate axis indicating the chromaticity component according to the ratio
Of the class number on the coordinate axis indicating the luminance component
Determining the ratio is a means for solving the above-mentioned problem.

Further, in the limited color determination method of the present invention, a representative color is obtained by performing a discriminant analysis on the frequency distribution of each pixel value of the image data represented in the uniform color space, and the obtained representative color is obtained. The frequency distribution in the uniform color space is obtained again, a smaller number of representative colors are obtained by performing discriminant analysis on this frequency distribution, and the representative colors are gradually reduced by repeating this process. To determine the frequency distribution at each stage or at any stage to reduce the number of classes in the coordinate axis showing the luminance component,
Changing the ratio with the class number on the coordinate axis indicating the chromaticity component is a means for solving the above problem.

The limited color determination method of the present invention reads the lower bits of digital data representing the values of the three primary colors of each pixel of the color image data, and determines the periodicity at which the same value appears in the lower bits. On the other hand, for each pixel value of the color image data expressed in a coordinate system of a uniform color space, a frequency distribution on a coordinate axis indicating a luminance component and a coordinate axis indicating a chromaticity component in the uniform color space is obtained, and the frequency distribution is calculated. The weight distribution is corrected based on the periodicity data, and the frequency distribution is corrected by differentiating the class distribution on the coordinate axis indicating luminance and the class number on the coordinate axis indicating chromaticity with respect to the frequency distribution with the weight correction, and discriminant analysis. Is the means for solving the above problem.

The limited color determining apparatus according to the present invention is a limited color determining apparatus which approximates color image data represented by a coordinate system of a uniform color space with a limited representative color. So that the rank on the coordinate axis indicating the component and the rank on the coordinate axis indicating the chromaticity component are different,
A class number setting means for setting the number of classes of each coordinate axis, wherein each of
A three-dimensional division based on the set class of coordinate axes
Frequency distribution calculating means for obtaining a frequency distribution of the color image data for each raster; and performing a discriminant analysis on the frequency distribution of each cluster to divide the uniform color space.
Color region dividing means, and each color component of the divided color region
And a representative color determining means for determining a representative color in each color area based on the color information .

Further, the limited color determining apparatus of the present invention obtains a representative color by performing a discriminant analysis on a frequency distribution of each pixel value of color image data represented by a coordinate system of a uniform color space, thereby obtaining the representative color. A periodicity calculating means for approximating the image data with limited representative colors, and a periodicity calculating means for obtaining a periodicity coefficient in which the same value appears in lower bits with respect to digital data representing three primary colors of the color image data; A frequency distribution calculating means for calculating a frequency part of each pixel value of the color image data with respect to a coordinate axis indicating a luminance component and a coordinate axis indicating a chromaticity component in a color space; and the frequency distribution calculated by the frequency distribution calculating means is defined by the periodicity. Frequency distribution correction means for performing weight correction using a coefficient; and a coordinate system indicating a class number of a coordinate axis indicating a luminance component and a chromaticity component with respect to the weight corrected frequency distribution. The frequency distribution by correcting the number of classes of the shaft as a different, that is formed by and a representative color determining means for determining a representative color by performing a discriminant analysis was solutions of the problems.

[0019]

According to the configuration of the first aspect, the image data converted into the uniform color space is subjected to the discriminant analysis with respect to the frequency distribution of the pixels for each coordinate axis, so that the uniform color is obtained.
When dividing a space to obtain a representative color in each color region, by making the rank of the coordinate axis indicating luminance and the rank of the coordinate axis indicating chromaticity different, the luminance or the color when selecting the representative color is determined. The degree can be emphasized. That is, in consideration of the color tone of the image data and the fact that human vision is sensitive to luminance, the selected representative color may be set to have a large change on the luminance side or may have a large change on the chromaticity side. It is possible to do.

According to the second aspect of the present invention, a representative color whose luminance is emphasized is selected in accordance with a human visual characteristic which is sensitive to luminance, and a representative color corresponding to human vision is selected. Color can be selected. For example, when only a small number of representative colors are selected in a portion where the color of the image data changes slowly, a false contour occurs,
In particular, since human vision is sensitive to luminance, when a color with similar luminance is selected as the small number of representative colors, false contours will be emphasized. That is, by selecting a representative color rich in change in luminance, it is possible to prevent false contours from being emphasized.

An image in which false contours are easily emphasized as described above is one in which the color distribution is concentrated on the same hue and has various luminance components. In particular, false contours can be prevented by selecting a representative color whose luminance is emphasized.

According to the third aspect of the present invention, when the color distribution of image data is dispersed in various hues and concentrated on a relatively small number of luminance components, human vision is reduced. Even if it is sensitive to luminance, false contours will not be emphasized.Therefore, a representative color with enhanced chromaticity, that is, a representative color of various hues can be selected, so that a representative color can be displayed. The resulting image can be close to the original image data.

According to the fourth aspect of the present invention, when a representative color is selected by discriminant analysis, the color is not reduced to the required representative color at a time, but, for example, the equalization of the original image data is performed. The frequency distribution in the color space is determined, the representative color is determined to be 10,000 colors by discriminant analysis, the frequency distribution in the uniform color space of the representative color is determined, and the representative color is reduced to 1000 colors by discriminant analysis. To do. At each stage, discriminant analysis is performed by emphasizing the luminance or chromaticity corresponding to the color distribution of the original image data or the representative color of each stage, thereby providing a statistical analysis of the color distribution of the original image data. It is possible to select a representative color suitable for the property.

In other words, the process simply repeats the non-adaptive method of selecting a representative color by discriminant analysis, and corresponds to the statistical properties such as the color distribution of image data in the same manner as the adaptive method requiring complicated processing. The selected representative color can be selected.

Further, according to the configuration of the fifth aspect, in a portion where the color gradually changes in the image data, when the color data of the pixels arranged continuously is read in order, the lower order is set. When the value of the bit increases, it is carried up to the upper bit, and a series of cycles or the reverse cycle in which the value of the lower bit decreases at a stretch is repeated.

Therefore, each pixel successively arranged in a portion where the color change in the image data is gradual has a periodicity in the value of the lower bit, and the pixels arranged consecutively have a lower value. By examining the periodicity, it is possible to specify a portion where the color of the image data changes gradually and the color of that portion, and it is possible to estimate the degree of the gradual color change from the strength of the periodicity. .

Therefore, when performing a discriminant analysis by creating a histogram indicating the frequency of appearance of colors in the image data, a portion of the image data having a gradual change in color is specified and the period of the lower bits is determined. For the color of the gradual change in color having the characteristic, a weighting coefficient is determined based on the strength of the periodicity, and the histogram is weighted using the weighting coefficient to perform the discriminant analysis. The number of representative colors to be displayed can be increased, and a false contour generated by displaying a portion having a gradual change in color with a small representative color is suppressed.

According to the limited color determining apparatus of the present invention, the rank setting means sets the rank of the coordinate axis indicating the luminance component and the rank of the chromaticity component in the uniform color space to be different. Set by the frequency distribution calculation means,
Divide three-dimensionally by the set class of each coordinate axis
Frequency distribution of the color image data for each cluster
Is required. Then, by the color area dividing means,
Uniform color by performing discriminant analysis on the selected frequency distribution
The space is divided and based on each color component of the divided color area.
Since the representative color in each color area is determined by the representative color determining means, the selected representative color is changed to the luminance side in consideration of the color tone of the image data and the sensitivity of human vision to luminance. It is possible to increase the number of colors or change the number of colors on the chromaticity side.

Further, according to the limited color determining apparatus of the seventh aspect, the periodicity calculating means obtains a periodicity coefficient indicating the gentleness of the color change, and the frequency distribution correcting means calculates the frequency distribution. After the frequency distribution obtained by the means is weighted and corrected by the periodicity coefficient, the representative color determining means performs the discriminant analysis, so that many representative colors are selected from the portion where the color changes gradually. Will be. In addition, the frequency distribution correction unit corrects the frequency distribution by making the rank of the coordinate axis indicating the luminance component and the rank of the coordinate axis indicating the chromaticity component different, so that the gradual change of the color tone and color of the image data is considered. The selected representative color is selected. In addition, billing
According to the limited color determination method described in claim 8, the limited color is determined according to claim 1.
In the constant color determination method, for each coordinate axis in a uniform color space
The variance of the pixel distribution is calculated in
In the coordinate axis indicating the chromaticity component according to the variance ratio of
Of the class number on the coordinate axis indicating the luminance component for the class number
The ratio is determined.

[0030]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a schematic diagram of a limited color determining apparatus according to the present embodiment, and FIGS. 2 to 8 are diagrams for explaining a limited color determining method according to the present embodiment. Hereinafter, the limited color determining apparatus of the present embodiment will be described with reference to FIG.

The above-described limited color determining apparatus includes an image storage unit 1 for storing image data, a lower 3 bit storage unit 2 for storing lower 3 bits of the image data, and each pixel in the lower 3 bits of image data. And the periodicity calculating means 3 for reading the values of the lower 3 bits of the surrounding pixels along the radiation direction, determining the periodicity of the lower 3 bits of the read pixels, and normalizing the periodicity. A periodicity coefficient storage means 4 for storing the periodicity normalized by the arithmetic means 3 as a coefficient; and a color conversion means 5 for converting the image data composed of each value of RGB into a value in L ^* a ^* b ^* space. And the L ^* a ^* b ^* space is subdivided into a number of clusters based on the number of classes used to determine the frequency distribution of each pixel on each of the L ^* , a ^* , and b ^* coordinate axes, and these clusters are assigned cluster numbers. L The value of the pixel converted to a value in the ^* a ^* b ^* space is
A cluster image storage unit 6 for storing the cluster numbers of the clusters including the pixel in place of the cluster number; a histogram calculation unit 7 for obtaining the frequency distribution of the pixel for each cluster for each coordinate axis in the L ^* a ^* b ^* space; Histogram storing means 8 for storing the obtained histogram, average and variance calculating means 9 for obtaining the average color and variance from the histogram, and average and variance storing means 10 for storing the obtained average and variance.
A first spindle calculating means 11 for obtaining a first spindle, a first spindle storing means 12 for storing the obtained first spindle, a weighted variance calculating means 13 for obtaining a weighted variance, and a division threshold value by discriminant analysis. , A division threshold storage unit 15 for storing the obtained division threshold,
A color region dividing unit 16 that divides a color region by a division threshold according to the weighted variance and the first principal axis, and until the color region that can be divided disappears or the number of divided color regions reaches a predetermined number of display colors, Division control means 17 for controlling the processing by each means repeatedly; and a coordinate axis indicating a luminance component L ^* according to a ratio of dispersion to each coordinate axis in L ^* a ^* b ^* space before dividing into each color area. Emphasis coefficient designating means 18 for designating how many times (1/8 to 32 times) the class number is on the coordinate axis indicating the chromaticity components a ^* b ^* ; Coefficient calculating means 19 for calculating the rank of each coordinate axis by the magnification (emphasis coefficient) designated by
Weighting the histogram read from the histogram storage means 8 by the emphasis coefficient storage means 20 for storing the rank of each coordinate axis obtained by the above, and the periodicity coefficient read from the periodicity coefficient storage means 4, and A histogram updating means 21 for changing the class number of the histogram to the class number stored in the emphasis coefficient storage means 20 and rewriting the histogram, and storing the rewritten histogram in the histogram storage means;
The inverse color conversion means 22 for reading out the limited color obtained by performing the operation 1 from the averaging / dispersion storage means and performing the inverse conversion of the color conversion means 5 and the representative color assigning means 23 for replacing the color of each pixel with the limited color. A limited color pallet means 24 for sorting the limited colors by the distance in the color space and registering the representative colors in the color map 30a; and a screen on which the image data replaced with the representative colors is approximately displayed, A color correcting means 25 for enabling a partial color correction;
The repetition control means 26 for controlling the repetition of the operations of the means 2 to 21 for the limited color obtained by performing the operation 21 is mainly constituted.

The image storage means 1 has an image memory 10a and stores (R, G, B) values (each color component is 8 bits) of each pixel of the image data. The lower 3 bits storage unit 2 stores the lower 3 bits of the (R, G, B) value of each pixel of the image data stored in the image storage unit 1.

The periodicity calculating means 3 reads out the lower 3 bits of data of each pixel from the lower 3 bits storage means 2 and, on the image, surrounds each pixel in a vertical, horizontal, left diagonal, right diagonal radial manner. In the pixels arranged continuously in the direction, the number of times that the same value appears in the lower 3 bits in a cycle of 4 pixels or less is measured, and a cycle value described later is obtained for each pixel. Further, a ratio when the maximum value of the periodic value is set to 1 is obtained as a periodicity coefficient for each pixel.

The periodicity coefficient storage means 4 stores the periodicity coefficient for each pixel as an array similar to the pixels. The color conversion means 5 converts the RGB values of each pixel of the image data into coordinates in L ^* a ^* b ^* space. The cluster image storage means 6 divides the L ^* a ^* b ^* space into clusters according to the rank of each coordinate, and stores the values of the pixels included in each cluster by replacing them with cluster numbers. The histogram calculation unit 7 obtains a frequency distribution of each pixel belonging to each cluster, creates a histogram, and stores the histogram in the histogram storage unit 8.

The average / dispersion calculating means 9 finds the average and the variance of the pixel distribution for each coordinate axis in the L ^* a ^* b ^* space or the color region obtained by dividing the L ^* a ^* b ^* space by a threshold value. The first spindle calculating means 11 calculates the maximum coordinate axis of the variance read from the average / variance storage means 10 in the L ^* a ^* b ^* space or the color area obtained by dividing the L ^* a ^* b ^* space by a threshold value. The first main shaft. The first spindle storage means 12 stores the obtained first spindle.

The weighted variance calculating means 13 obtains a weighted variance by multiplying the variance by the relative frequency of the number of pixels included in the color area with respect to the total number of pixels. The division threshold value determining means 14 determines a division threshold value by the discriminant analysis. The division threshold storage unit 15 stores the division threshold obtained by the division threshold determination unit 14.

The color area dividing means 16 sorts the values of the weighted variances of the respective areas obtained by the weighted variance calculating means 13 to obtain the color area having the largest weighted variance. Is read out from the first spindle storage means 20, and the color area (color space) is bisected by the division threshold of the first spindle read out from the division threshold storage means 15. The division control unit 17 repeats the above-described calculation and division for each color region until the total number of color regions divided by the color region division unit 16 reaches the specified number of colors or the color region cannot be further divided. To control.

After the above division is completed, the representative color allocating means 23 calculates the average color of each color area as an average. The average color of the original image data is read out from the distributed storage unit 10 and the average color is adopted as a representative color of each color region. In addition, in the original image data, the average color of the color region is assigned to the pixels included in each color region. The pixel color is replaced with the representative color. The image data obtained by replacing the color of each pixel with the representative color is stored in the image memory 30d.

The emphasis coefficient designating means 18 calculates the ratio between the coordinate axes of the variance of each coordinate axis in the L * a * b * space obtained by the average / variance calculating means 9 and calculates the class number of each coordinate axis. 1/8 of other coordinate axes according to the ratio
Times, 1/4 times, 1/2 times, 2 times, 4 times, 8 times, 16 times,
Specify 32 times or specify the magnification. The emphasis coefficient calculating means 19 obtains the class number of each coordinate axis according to the magnification.

The emphasis coefficient storage means 20 stores the rank of each coordinate axis obtained by the emphasis coefficient calculation means 19. The histogram updating means 21 stores the periodicity coefficient storing means 4
Is added for each cluster including the corresponding pixel, and the sum of the periodicity coefficients for each cluster and the number of pixels included in each cluster are added. Then, the value obtained for each cluster as described above is newly set as a histogram.

Further, the histogram updating means 21 subdivides the L ^* a ^* b ^* space into new clusters different from the clusters according to the rank of each coordinate axis stored in the emphasis coefficient storage means 20. The histogram is changed to the number of pixels for each new sub-cluster.

That is, the L ^* a ^* b ^* space is divided into new clusters according to the magnification specified by the emphasis coefficient specifying means 18, and the histogram is updated based on the new clusters. Therefore, when the weighting correction is performed using the periodicity coefficient and the luminance component is emphasized at the magnification specified by the emphasis coefficient specifying unit 18, the L ^* a ^* b ^* space is divided into new clusters, And the histogram based on the new cluster is weighted and corrected by the sum of the periodicity coefficients of each new cluster. Then, the histogram updating unit 21 rewrites the histogram stored in the histogram storage unit 8 with the histogram updated as described above.

The limited color color pallet operation means 24 sorts the representative colors obtained by the operations of the above-described means in terms of the distance in the color space, and registers the representative colors as color codes in the color map 30a. The color correcting unit 25 changes the color of the display image composed of the limited colors obtained by the operations of the above units on the screen of the display 30b displaying this image.
This is used when you want to make partial changes.

Then, the color correcting means 25 surrounds a portion of the display image whose color is to be changed with a rectangle according to a mouse instruction, focuses on the color to be changed in the rectangle, clicks the mouse, and selects a color. As a result, on the color palette screen, the selected color portion is highlighted in reverse, and by selecting colors in order from the neighboring colors,
It is possible to replace colors in order from the closest color in the color space.

The limited color determination device having the above-described configuration displays full-color image data in a limited display color on the color map display device 30 by using the limited color determination method. In the present embodiment, the color map display device 30 is also capable of displaying image data stored in the image storage unit in full color, and comparing the full color image with the limited color image, The color correction means 28 can correct the color. Note that reference numeral 30c denotes the image memory 10a,
This is a D / A converter for displaying the image data of 30d on the display 30b.

Next, a limited color determining method using the above limited color determining apparatus will be described. (A) Measurement of the degree of color change of image data to be displayed In the conventional method, the degree of change of the color of image data, that is, the color of each pixel in an image and the color of pixels near the pixel are large. When measuring whether the colors are different or have substantially the same color, the measurement is often performed using the upper 5 bits of the RGB data of the image represented by each 8 bits of RGB.

However, information indicating the gradual change in color is as follows:
Rather, it is included in the lower three bits lost by this. That is, in image data with little color change,
Even if the upper bits of each pixel are compared, the degree of color change cannot be measured, and the portion where the color change is extremely gentle,
It cannot be distinguished from a portion where the color change is gentle to some extent. Therefore, in this embodiment, the lower three
By examining the periodicity of the bit values, the gradual change in color in the image data to be displayed is specified.

By the way, the values of the lower 3 bits are from 0 to 7
Limited to 8 values up to When the color change due to the difference in the position of each pixel of the image is gradual, the value of the data of each pixel continuously changes with a small variation amount between the pixels arranged continuously. Therefore, in a portion where the color change is gradual, a certain value range within the range of 0 to 7 of the lower 3 bits appears almost regularly. Therefore, in the present embodiment, as shown in FIG.
Vertical (upper / lower), horizontal (left / right), diagonal left (upper left / lower right), diagonal right (upper right / lower left) centering on each pixel 41 ...
The lower three bits (stored in the lower three bits storage means 2) of the pixels arranged continuously in the eight radial directions are read, and the above-mentioned regularity (periodicity) is obtained for each of the RGB color components.
Find out.

As shown in FIG. 3, a specific method of calculating the periodicity by the periodicity calculating means 3 is as follows.
Are read in the digital data of the pixels 41... While sequentially moving one pixel at a time along the above-mentioned direction until the same value as the value of the lower bits of the pixels 41 appears. If the number of pixels moved until the same value is first found is 4 or less, the lower 3 bits are read while sequentially moving to the next pixel until the next same value is found, and then the same value is found. If the number of moving pixels at the time is equal to the first moving pixel number, it is considered that the pixel of interest 41a has periodicity.

While the same value periodically appears with the same number of movements, the lower three bits of the color data of the pixels 41 are read while moving along the above-mentioned direction. Then, the number of appearances in which the same value appears with the same movement number is counted.
And when the same value stops appearing periodically,
The reading of the pixel data ends. The above operation is performed in all eight directions, and the total count of the number of appearances of the same value is defined as the cycle value of the pixel of interest 41a.
The above operation is performed with all the pixels 41 of the image data as the target pixel 41a. The measurement of such a period value is performed for each of the RGB color components.

The above operation is performed, for example, in the case shown in FIG.
When attention is paid to the pixel 41a on the image data, looking at the values of the pixels 41... Continuous in the horizontal (right in FIG. 2) direction from the pixel 41a, for example, a pixel 41b having the same value every three pixels May appear. In such a case, the number of appearances of the same value appearing in the cycle of three pixels is counted, and the same operation is performed for the remaining radiation directions to count the number of appearances of the same value. The period value of the pixel 41a is used.

Next, the cycle value of each color component is added for each pixel. Then, the sum of the cycle values for each pixel is calculated at a ratio where the peak is 1, and the sum is defined as a periodicity coefficient used for weighting described later. The periodicity factor b _i is obtained by the following equation. b _i = a _i / a _max Here, i represents each pixel, a _i represents the sum of the period values in each pixel, and a _max is the maximum period value of the sum a _i of each period value. a _i (maximum value).

The thus obtained periodicity coefficients b _i is stored in the periodicity coefficient storage unit 4. Then, the histogram updating unit 21, the periodicity factor b _i is allocated to the position of each pixel is converted to a uniform color space, and a state in which the periodicity factor b _i to the position of each pixel on the uniform color space is arranged Become.

When the histogram indicating the number of pixels included in each cluster is updated by the histogram updating unit 21 in the uniform color space divided into clusters, each of the pixels included in each cluster is updated. adding periodicity coefficients b _i, the sum of the periodicity coefficient b _i and the weighting coefficient of each cluster. Then, these weighting factors are added to the number of pixels of each cluster, respectively, and this is used as a new histogram. When performing discriminant analysis on the first main axis, the first histogram
The number of pixels for each class of the main axis, that is, the frequency distribution for the first main axis is obtained, the number of pixels included in each class of the first main axis is divided by the total number of pixels from the frequency distribution, and the occurrence probability of pixels is obtained.
This occurrence probability is used.

(B) Selection of representative color In order to select a representative color, first, the color distribution of the image data is arranged in a color space.
Use a uniform color space. As the uniform color space, the L ^*
Use u ^* v ^* or L ^* a ^* b ^* . Therefore, the value of each pixel of the image data composed of RGB data is represented by L ^* u ^* v ^* or L ^* u ^* v ^*.
Convert to ^* a ^* b ^* . In the present embodiment, as an example, the details of the conversion by the color conversion means 5 will be described using L ^* a ^* b ^* . Note that the uniform color space of the present invention is L ^* a ^* b
^* It is not limited to space.

Conversion from RGB data to L ^* a ^* b ^* is performed by the following equation via the XYZ system recommended by CIE 1931. X = 0.61R + 0.17G + 0.20B (1a) Y = 0.30R + 0.59G + 0.11B (1b) Z = 0.066G + 1.12B (1c)

When Y / Yn> 0.008856 L ^* = 116 (Y / Yn) ^1/3 −16 (2a) When Y / Yn ≦ 0.008856 L ^* = 903.3 (Y / Yn) ) ...... (2b)

X / Xn> 0.008856, Y / Yn>
A ^* = 500 [(X / Xn) ^1/3 − (Y / Yn) ^1/3 ] in the case of 0.008856, Z / Zn> 0.008856 (2c) b ^* = 200 [(Y / Yn ) ^1/3 − (Z / Zn) ^1/3 ] (2d)

0.008 for X / Xn, Y / Yn, and Z / Zn
In the case where there is a value of 856 or less, the equations (2c) and (2
d) (X / Xn) ^1/3 , (Y / Yn) ^1/3 , (Z / Zn)
^1/3 terms are 7.787 (X / Xn) +16/11
6,7.787 (Y / Yn) + 16 / 116,7.78
7 (Z / Zn) +16/116. here,
The chromaticities of R, G, B and white conform to the NTSC television standard.

(C) Frequency distribution of pixels in L ^* a ^* b ^* space The value of each pixel may deviate from an 8-bit integer value at the stage when RGB is converted to L ^* a ^* b ^*. become.
Therefore, in order to obtain the frequency distribution in the L ^* a ^* b ^* space, the rank of each coordinate axis is determined, the L ^* a ^* b ^* space is divided into clusters by the rank of each coordinate axis, and the pixels included in the cluster are determined. You need to find the number.

To divide the L ^* a ^* b ^* space into a required number of colors and select an average color of each color region as a representative color, Watanabe's algorithm (IEICE Transactions D Vol.
J70-D No. 4 pp. 720-726 198
(April 1995) with improvements. The main improvement is that instead of using the upper 5 bits of RGB as the original data, R
A value converted from an 8-bit GB value to L ^* a ^* b ^* was used. Thus, the 8-bit precision of the data can be used as it is.

The class number of each coordinate axis in the initial space to be divided is changed from, for example, RGB 32 × 32 × 32 to L ^* a ^* b ^*.
The rank of the coordinate axis indicating the luminance component L ^* is quadrupled to the rank of the coordinate axis indicating the chromaticity component a ^* b ^* , such as 256 × 64 × 64, so that the luminance component L ^* is emphasized. Was.
This is the report of Kira et al. (IEICE, IE83-92,
"Adaptive natural color image display with a limited number of representative colors"
(Kira, Inoue, Fukui) because human visual characteristics are about four times more sensitive to luminance distribution than chromaticity components.

More specifically, the input image data usually has RGB values arranged in pixel order. Therefore,
Considering an RGB space having coordinates of RGB values, RGB
The space is a three-dimensional space in which R, G, and B are coordinate axes, and each coordinate axis has an 8-bit value, that is, a value of 0 to 255. Therefore, a histogram of each color component is obtained by associating the number of pixels having the RGB value with a point in space.

That is, histogram (r, g, b) = RGB (r,
g, b) is the total number of pixels. Next, the uniform color space L ^* a ^* b ^*
Is obtained by converting the value (r, g, b) of each pixel represented by RGB into (L ^* a ^* b ^* ) by the above equation. That is, histogram (L ^* a ^* b ^* ) = L ^* a ^* b ^*
It is the total number of pixels of (L ^* a ^* b ^* ). However, the value of RGB is
Since each is 8-bit digital data, the value is an integer value of 0 to 255, but the value of (L * a * b *) is 0
Since it deviates from an integer value of 255 and emphasizes the luminance component, it is necessary to provide an appropriate class number.

In this embodiment, in order to enhance the luminance, that is, the L ^* component as described above, and to prevent the data capacity from increasing due to the conversion to L ^* a ^* b ^* , each L
^It is necessary to determine the class interval of ^* a ^* b ^* . As the data capacity, since each value of RGB is set to 8 bits, the maximum is set to 8 bits. Therefore, the emphasized luminance L ^* takes a value of 0 to 255. Therefore, the value of the L ^* component of the pixel obtained by the above equation is assigned to 0 to 255 corresponding to the size. Also, a
Regarding the ^* and b ^* components, a ^* , b ^* : L ^{* is set} to 1: 4 in order to emphasize the luminance component four times. That is, a ^* , b ^*
The component has a value of 6 bits, 0 to 63.

Here, the actual calculation of the class interval is performed for each L ^*
When a ^* b ^* is 8-bit or less data, it is obtained by the following equation. L ^* interval = (L ^* max-L ^* min) / (256 / m) a ^* interval = (a ^* max-a ^* min) / (256 / n) b ^* interval = (b ^* max-b ^* min) ) / (256 / n) L ^* interval, a ^* interval, b ^* interval are L ^* ,
^Indicates the class interval of a ^* and b ^* . L ^* max, a ^* max, b ^* max
^Indicates the maximum value of L ^* , a ^* , and b ^* , respectively. L ^* min, a ^*
min and b ^* min indicate the minimum values of L ^* , a ^* and b ^* , respectively. Further, m and n are changed depending on the ratio between the luminance component (L ^* ) and the chromaticity components (a ^* , b ^* ). In this embodiment, L ^* : a ^* , b ^{* is changed} to 4: In order to set it to 1, m of the luminance component is set to 1 and n of the chromaticity component is set to 4.

Then, the L ^* a ^* b ^* space in which the class interval is determined as described above is three-dimensionally divided by the class of each coordinate axis, and the divided region is defined as a color cluster. Then, a cluster number is assigned to each cluster. And
The value of each pixel is replaced with the cluster number of the cluster including that pixel, and this is stored in the cluster image storage unit 6. Then, the histogram calculation means 7 obtains a frequency distribution for each cluster of the image data.

Referring to FIGS. 4 and 5, a brief description will be given of a case where the rank of the coordinate axis indicating luminance is larger than the rank of the coordinate axis indicating chromaticity, and a representative color is selected by discriminant analysis. 4 and 5, a description will be given using a two-dimensional coordinate system with one coordinate axis indicating chromaticity in a uniform color space.

First, when the y-axis is luminance and the x-axis is chromaticity, and each class is 8, the color space is divided by discriminant analysis as shown in FIG. And
When the class number on the x-axis is set to 4 and the class number on the y-axis is twice the class number on the x-axis, as shown in FIG. Cannot be divided, and the number of divisions of the luminance axis increases accordingly. That is, when the average color in the divided color space is set as the representative color, many colors having the same hue but different luminances are selected as the representative colors, and colors having different hues are not selected so much. Will be selected.

(D) Enhancement of Luminance or Chromaticity In this embodiment, in the space division method, the sampling of the luminance component is emphasized to emphasize a portion where the change of the luminance component is gradual. A method of identifying a portion and a color with a gradual change in luminance by focusing on that the division with as the first main axis is a main division, and further giving sampling in which the luminance component is emphasized for the chromaticity component It is.

Specifically, L ^* a ^* b ^* (or L ^* u)
^* v ^* ), 16 in the luminance direction, such as 256 × 16 × 16
The sample is emphasized twice, and the number of limited colors is set to, for example, 10
The representative color is designated by a smaller number such as 0 color, and the representative color is obtained by the division method of (b). Thus, in this case, 100 colors emphasized 16 times are selected for the luminance component. These are the top 100 colors that emphasize the colors of the portions where the luminance changes gradually.

Then, as described above, the emphasis coefficient designating means 18 calculates the ratio between the coordinate axes with respect to the variance in the direction of each coordinate axis of the uniform color space obtained by the average / variance calculating means 9, and corresponds to the ratio. Then, the magnification (emphasis coefficient) of the rank of the coordinate axis indicating the luminance with respect to the rank of the coordinate axis indicating the chromaticity is 1/8, 1/4, 1/2, 1 or 2 times.
, 4 times, 8 times, 16 times, 32 times, 64 times.

The emphasis coefficient calculating means 19 determines the class of each coordinate axis based on the magnification determined as described above. For example, when the magnification of the chromaticity component a ^* b ^* with respect to the luminance component L ^* is doubled, the rank of the coordinate axis of each L ^* a ^* b ^* is 256 × 128 × 128. Classification interval determined by the class number
Remembers.

(E) Change of histogram The histogram coefficient is stored in the histogram storage unit 8 using the weighting coefficient for each pixel stored in the periodic coefficient storage unit 4 and the class stored in the enhancement coefficient storage unit 19. Update the histogram.

First, updating of the histogram using the periodicity coefficient stored in the periodicity coefficient storage means 4 will be described. First, the periodicity coefficient stored for each pixel in the periodicity coefficient storage means 4 is arranged at the position of each pixel converted into a uniform color space, and the periodicity coefficient of each pixel included in each cluster is stored in the cluster. Add each time. Then, a histogram is created using the value obtained by adding the sum of the periodicity coefficients added for each cluster and the number of pixels for each cluster as the number of pixels for each cluster, and this is used as a new histogram.

That is, as shown in FIG. 6 and FIG.
(Actually, a histogram in a uniform color space, that is, a three-dimensional space, but here, a one-dimensional histogram is used for ease of explanation). By adding a histogram representing the sum for each cluster, a weight-corrected histogram 45 shown in FIG. 8 can be obtained, and using this, a representative color is selected by discriminant analysis described later.

The histogram updating means 21 reconstructs a histogram by using the class number stored in the emphasis coefficient storage means 20. In the histogram calculation means 7, a histogram is created with a class number in which the chromaticity component is enhanced four times with respect to the luminance component regardless of the color tone of the image data. Decide.

Then, the uniform color space is re-divided into new clusters of a new class, and the number of pixels for each new cluster is set as a new histogram. When weighting is performed together with the periodicity coefficient, the sum of the periodicity coefficients for each new cluster is used as a weighting coefficient, and the weighting coefficient is added to the number of pixels for each new cluster to form a new histogram. Then, the histogram stored in the histogram storage unit 8 is updated to a new histogram obtained as described above.

(F) Selection of representative color by discriminant analysis method The selection of a representative color by the discriminant analysis method of the present embodiment is basically the same as that of the above-described prior art, and a threshold value is determined by using a histogram for each color component. And the color space is sequentially divided into a number of color regions at this threshold value. In the present embodiment, the histogram weighted by the periodicity and reconstructed by the emphasized class number is used as the histogram. And perform discriminant analysis.

Here, a method of determining the threshold T _i by discriminant analysis will be described. Assuming that the number of pixels in the area S is N and the number of pixels in any rank i on one coordinate axis is n _i , the occurrence probability q _i
Is represented by the following equation (Equation 1). The number of pixels at this time is obtained from the histogram updated by the histogram updating unit 21 as described above.

That is, the value obtained by adding the weighting coefficient to the number of pixels of the new cluster divided by the new number of classes is used as the number of pixels for each cluster, and the number of pixels for each cluster included in an arbitrary class on one coordinate axis is calculated. The sum is the number of pixels n _i
Used as

(Equation 1) Then, the section [a, b] is divided into two sections C ₁ [a, k] and C ₂ [k + by a value k (a value in units of class) on the coordinate axis.
1, b], the pixel occurrence probabilities ω ₁ and ω ₂ in both sections, the average values μ ₁ and μ ₂ , and the variances σ ₁ ² and σ
In case of a ₂ ^2, define the value k of the coordinate axis as weighted variance sum sigma _w ² given by: (Equation 2) is minimized.

(Equation 2)

Minimizing the weighted variance sum σ _w ² is equivalent to maximizing the variance σ _B ² between the two sections given by the following equation (Equation 3).

(Equation 3)

K ^* that maximizes σ _B ² (that is, the threshold value T _{i to be obtained} ) can be obtained sequentially as in the following equation (Equation 4).

(Equation 4) Here, μ _T is the total average and is represented by the following equation (Equation 5).

(Equation 5)

As described above, the threshold value T _i of each color component can be determined. Then, the color space is firstly divided into two regions by the obtained threshold value T _i , and the coordinate axis of the color component having the largest variance σ ² among the respective color components is defined as the first principal axis. , The threshold value T _{i of the} color component
The color space is divided into two color regions on a division plane orthogonal to the first principal axis at the position of the first principal axis corresponding to.

Next, a weighted variance sum σ _w ² of each of the divided color regions is obtained, these are compared between the color regions, and the color region having the largest weighted variance sum σ _w ² is divided next. Area. Then, in the color region, a threshold value _Ti and a first principal axis are obtained as if the color space were divided, and the color region is further divided into two color regions. Next, a color region having the largest weighted variance sum σ _w ² is obtained in all the divided color regions, and the obtained color region is divided into two as described above. Then, when the number of the divided color areas becomes equal to the number of the representative colors to be displayed, the division is completed. Next, the average value of each color component in each color area is converted from L ^* a ^* b ^* to RGB by the inverse color conversion means 22 to obtain a representative color.

Then, this representative color is registered as a color code in the color map 30a of the color map display device 30, and the color distribution in the color space of the color image data to be displayed is made to correspond to each of the color regions, and each of the image data is A color similar to the original color of the pixel is assigned to the pixel as a color code from the color map 30a. In this way, the image data in which the color is assigned to each pixel and the color is limited is displayed on the display 30b of the color map display device 30 having the color map 30a in which the color code is registered.

The image data of the limited color can be stored in a storage device or the like, and the image processed once can be instantaneously displayed on the display 30b, and the limited image data can be displayed. Since the memory capacity of the data is smaller than that of the image data of the original natural image, it can be easily handled even by a device having a small memory capacity.

(G) Determination of a representative color by repeating the above operation. In the limited color determination method of the present embodiment, the representative colors selected as described above are repeatedly selected for the selected representative colors, so that the representative colors are gradually reduced. That is, the repetition control means 17 selects the representative color by the above-described operation again using the image represented by the representative color selected as described above as the original image data.
This is repeated a set number of times to reduce the number of representative colors finally required.

Each time the representative color is repeatedly selected, the histogram is corrected by the periodicity coefficient calculating means 3 and the enhancement coefficient calculating means 19, and the histogram is updated by the histogram updating means. The parts are weighted and the luminance or chromaticity is emphasized. Therefore, the representative color is narrowed down in the image data while adjusting the statistical properties of the color distribution and the luminance distribution, and the representative color suitable for the statistical property of the image data is adjusted similarly to the adaptive method. The color will be selected.

As described above, the limited color determination device of the present embodiment specifies the portion where the color changes gradually, based on the periodicity of the lower three bits of the image data, and uses the periodicity as a coefficient to determine the color distribution of the image. Weighting correction, select many colors from the gradual part of the color change, and make the number of ranks of the coordinate axis indicating the luminance greater than the number of ranks of the coordinate indicating the chromaticity, in response to the characteristics of human vision By selecting many colors with different brightness,
False contour can be prevented.

The variances of the luminance and the chromaticity in each coordinate axis are compared.
Further, by emphasizing the luminance, and by emphasizing the chromaticity of an image having a small change in luminance and many hues, it is possible to select a representative color suitable for the characteristics of the image data.
Further, by repeating the above operation and gradually reducing the representative colors, the measurement of the periodicity of the lower three bits and the comparison of the variance of the luminance and chromaticity on each coordinate axis are repeated, and the histogram corresponding to these is compared. Is repeated, and the representative colors can be narrowed down to those suitable for the statistical properties of the image data.

Note that the entirety of these processes can be configured as a program on a general-purpose computer and executed. Also, the memory capacity per pixel of the original image data does not necessarily need to be 24 bits, and the lower bits used for specifying the color change need not be the lower 3 bits. The number of lower bits to be taken may be changed according to the memory capacity per pixel of the original image data and the like. In addition, although it takes some processing time to reduce the number of colors according to the present invention, when displaying a compressed image obtained by this method, decoding processing is not necessary, so that it does not depend on a machine or a system, and is instantaneous. It is possible to display a compressed image with limited colors. For these reasons, the limited color determination method can be expected to be used in many industrial fields.

[0093]

According to the above-described method and apparatus for determining a limited color according to the present invention, a natural image which should be displayed on a full-color display apparatus can be displayed on a color map display apparatus which can simultaneously display only a limited number of limited colors. Can be displayed without.

That is, when the representative color is determined by the discriminant analysis, the class number on the coordinate axis indicating the luminance is different from the class number on the coordinate axis indicating the chromaticity, so that the distribution of the luminance and the chromaticity of the image is determined. Accordingly, a representative color whose luminance is emphasized or a representative color whose chromaticity is emphasized can be selected. In particular, by selecting a representative color whose luminance is enhanced, it is possible to select a representative color corresponding to a human visual characteristic that is sensitive to the luminance, and to suppress the false contour from being enhanced.

Further, the gradual color portion is specified by the periodicity of the lower bits, the histogram is corrected, and the number of representative colors selected from the gradual color portion is increased. Can be suppressed. Therefore, 1
C that requires a particularly small number of colors, such as 00 colors or less
Even for natural images such as D-ROMs, it is possible to effectively cope with the problem by identifying portions where conspicuousness is conspicuous, and obtaining a limited number of specified colors with the least discomfort among them. It is very effective in compression of.

[Brief description of the drawings]

FIG. 1 is a schematic diagram showing a limited color determination device of the embodiment.

FIG. 2 is a diagram for explaining a limited color determining method of the embodiment.

FIG. 3 is a diagram for explaining the above-described limited color determining method.

FIG. 4 is a drawing showing a coordinate system for explaining the above-described limited color determining method.

FIG. 5 is a drawing showing a coordinate system for explaining the above-described limited color determining method.

FIG. 6 is a diagram showing a histogram for explaining the above-described limited color determining method.

FIG. 7 is a diagram showing a graph for explaining the above-described limited color determining method.

FIG. 8 is a drawing showing a histogram for explaining the above-described limited color determining method.

[Explanation of symbols]

 2 Lower 3 bits storage means 3 Periodic calculation means 5 Color conversion means 7 Histogram calculation means 9 Average and variance calculation means 11 First spindle calculation means 13 Weighted variance calculation means 14 Division threshold value determination means 16 Color area division means 17 Division control Means 18 Emphasis coefficient designating means 19 Emphasis coefficient calculating means 21 Histogram updating means 22 Inverted color conversion means 23 Representative color allocating means 41, 41a, 41b Pixel 43 Histogram showing frequency distribution 44 Graph showing weighting coefficient 45 Weighted histogram

Claims (8)

(57) [Claims] 1. A class number of each coordinate axis is set in a uniform color space such that a class number on a coordinate axis indicating a luminance component is different from a class number on a coordinate axis indicating a chromaticity component , and the setting of each coordinate axis is set. Three-dimensionally divided by the class
For each cluster represented by the coordinate system of the uniform color space
Obtains a frequency distribution of the color image data, the said uniform color space is divided by performing determine by analysis to the frequency distribution for each cluster, put in each color region based on each color component of the divided color areas
That determine the representative color, the by representative colors determined, limited color determination method characterized by approximating the color image data. 2. When obtaining the frequency distribution in the uniform color space, the number of ranks of the coordinate axis indicating the luminance component is set to be larger than the number of ranks of the coordinate axis indicating the chromaticity component, and a representative color whose luminance is emphasized is selected. 2. The method for determining a limited color according to claim 1, wherein: 3. When calculating the frequency distribution in the uniform color space, a rank of a coordinate axis indicating a chromaticity component is set to be larger than a rank of a coordinate axis indicating a luminance component, and a representative color in which chromaticity is emphasized is selected. 2. The method for determining a limited color according to claim 1, wherein: 4. A representative color is obtained by performing a discriminant analysis on a frequency distribution of each pixel value of the image data represented in the uniform color space, and the obtained representative color is again subjected to a frequency distribution in the uniform color space. , And a smaller number of representative colors is obtained by performing discriminant analysis on the frequency distribution, and by repeating this, the number of representative colors is reduced step by step. 4. The method according to claim 1, further comprising: changing a ratio between a class number on a coordinate axis indicating a luminance component and a class number on a coordinate axis indicating a chromaticity component. Limited color determination method. 5. Reading the lower bits of digital data representing the values of the three primary colors of each pixel of the color image data, and determining the periodicity in which the same value appears in the lower bits. For each pixel value represented by the coordinate system, a frequency distribution on a coordinate axis indicating a luminance component and a coordinate axis indicating a chromaticity component in the uniform color space is obtained, and the frequency distribution is calculated based on the periodicity data. And performing a discriminant analysis by correcting the frequency distribution by making the class number on the coordinate axis indicating luminance and the class number on the coordinate axis indicating chromaticity different from each other with respect to the weight distribution corrected by weighting. Claim 1 or Claim 2 or Claim 3
Alternatively, the limited color determining method according to claim 4. 6. A limited color determination device that approximates color image data represented by a coordinate system of a uniform color space with a limited representative color, comprising: a class number on a coordinate axis indicating a luminance component of the uniform color space; Class number setting means for setting the class number of each coordinate axis so that the class number on the coordinate axis indicating the chromaticity component is different , and three-dimensionally dividing by the set class of each coordinate axis
Each comprising a cluster, a frequency distribution calculating means for calculating a frequency distribution of the color image data, a color area division unit for dividing the uniform color space by performing discriminant analysis the relative frequency distribution of each cluster, the Based on each color component of the divided color area,
And a representative color determining means for determining a representative color to be determined. 7. A representative color is obtained by performing a discriminant analysis on a frequency distribution of each pixel value of color image data represented by a coordinate system of a uniform color space, and the color image data is defined by a limited representative color. An approximate limited color determination device, comprising: periodicity calculating means for determining a periodicity coefficient in which the same value appears in lower bits of digital data representing three primary colors of the color image data; and a luminance component in the uniform color space. And a frequency distribution calculating means for calculating a frequency part of each pixel value of the color image data with respect to a coordinate axis indicating the chromaticity component and a coordinate axis indicating the chromaticity component. The frequency distribution calculated by the frequency distribution calculating means is weighted and corrected by the periodicity coefficient. A frequency distribution correcting unit, wherein the weight distribution of the weighted corrected frequency distribution is different from the rank of the coordinate axis indicating the luminance component and the rank of the coordinate axis indicating the chromaticity component. The frequency distribution is corrected as shall, discriminant analysis limited color determination apparatus characterized by comprising anda representative color determining means for determining a representative color by performing. 8. An image for each coordinate axis in the uniform color space.
Find the variance of the prime distribution, The chromaticity component is determined according to the calculated ratio of the variance between the coordinate axes.
The luminance component with respect to the class number on the coordinate axis indicating the minute
Determine the ratio of classes in the coordinate axis Specially
The method for determining a limited color according to claim 1, wherein: