JP2010011299A - Image processor and image processing method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain reproduction of colors that a human being would feel preferable, by making an effective use of a color gamut. <P>SOLUTION: An image processor includes a maximum saturation point estimating portion for acquiring not only a contour to the extent pixel values are distributed on the subject hue surface expressed by the hue value for representing each interval, but also a maximum saturation point which the value that the saturation acquires becomes a maximum, with respect to each of the intervals where a color space expressed by the hue and saturation, as well as, the intensity is divided as to the hue; a reference line setting portion for setting the reference line such that the intensity increases as the saturation increases passing through the maximum saturation point and intensity axis of the color space; and a correction direction calculating portion for acquiring the correction direction of each pixel value, according to the distance between a point of the color space for representing each pixel value of the input image and the reference line of the intervals to which each pixel belongs. Furthermore, the image processor includes a correction portion for acquiring the corrected pixel value, by correcting each pixel value according to the distance from the point of the color space for representing each pixel value of the input image to the contour toward the correction direction, and an output portion for outputting the corrected pixel value. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to an image processing apparatus and an image processing method for processing a color image. For example, the present invention relates to a technique for converting a color image signal input from the outside so as to achieve preferable color reproduction in a display device.

  There are various characteristics of the image equipment, and even if the same color image signal is input, the image equipment displays in various colors. This is due to the fact that each image device is not generally designed to output the same signal in the same color. Such a phenomenon becomes a serious problem when the color reproduction area of the output system, that is, the color gamut is narrower than the color gamut of the input system. For example, when an image displayed on a monitor of a PC (Personal Computer) or the like is displayed on an output system such as a printer, how to display colors existing in a range that cannot be reproduced becomes a problem.

When a color out of the color gamut occurs in this way, it is necessary to compress the color gamut of the image data, which has been addressed for a long time. Color gamut compression is basically clipping colors outside the gamut to the outermost gamut, but many colors are compressed into the same color and saturated, resulting in poor gradation. Resulting in. Therefore, linear compression, nonlinear compression, or the like is used. For example, Patent Document 1 describes a method of compressing the saturation toward the lightness axis while fixing the hue. Further, Patent Document 2 describes a method of compressing toward the middle point of the brightness axis of a predetermined color space with the same hue.
JP-A-3-73672 Japanese Patent Laid-Open No. 4-196675

  By the way, with the development of image equipment, the number of cases where the output color gamut is wider than the input color gamut has increased. In such a case, it is desired to reproduce the image more preferably by effectively using the color gamut of the output image device. In general, humans tend to prefer vivid images. Further, there is a similar demand when the input color image signal exists only in a narrow range close to the center of the color gamut.

  For such an image device, it is necessary to expand the color gamut instead of the conventional color gamut compression. Therefore, it can be considered that the conventional technology is adapted to expansion. An example of extending the saturation is shown in FIG. In this case, since only the saturation changes, the brightness does not change, and the effect of color conversion does not appear much. In order to achieve more preferable color reproduction, it is necessary to appropriately convert not only saturation but also lightness.

  As another extension example, an example of extending around the middle point of the lightness axis is shown in FIG. In this case, the color located below the midpoint increases the saturation but decreases the lightness. In the case of a medium lightness and medium saturation color, the color becomes dark and dull, and it cannot be said that color reproduction can be preferably performed.

  Further, as a problem when expanding the color gamut, the human visual color gamut and the device color gamut are not the same shape for each hue. For example, consider a case where the standard sRGB color space of a monitor is converted to CIELAB in a uniform color space. FIG. 13 is a chromaticity diagram obtained by converting the sRGB color space into CIELAB of the uniform color space and projecting it on the a * b * plane. CIELAB is a rectangular coordinate system, and can be converted into CIE Ch that can express a color with lightness, saturation, and hue by simply converting into a cylindrical coordinate system. 14A to 14C correspond to cross sections cut along the contour direction from the center in CIELAB in FIG. Specifically, FIG. 14A shows a gamut shape in a blue hue, FIG. 14B shows a gamut shape in a red hue, and FIG. 14C shows a gamut shape in a yellow hue. As shown in FIG. 14C, the vertex of the yellow outermost contour is located in a place where the lightness is high (bright). However, in the case of the blue hue in FIG. 14A, the vertex of the outermost contour is the lightness. Is located in a low (dark) place. In the case of red in FIG. 14B, the vertex of the outermost contour is located at an intermediate brightness level. Thus, the shape of the color gamut varies depending on the hue.

  Considering the shape of such a color gamut, it may be possible to extend the color toward the outermost vertex as shown in FIG. 15 for each hue, but when the lightness is higher than the vertex, the saturation increases. However, the brightness becomes low and the image becomes dark. In the case of a yellow hue (see FIG. 14 (c)), since the vertex coincides with the maximum brightness, there is no problem that the brightness is lowered, but the red hue (see FIG. 14 (b)), and In the case of a blue hue (see FIG. 14A), the lightness is significantly lowered, and a preferable color reproduction cannot be obtained as in the example shown in FIG.

  The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an image processing apparatus and an image processing method that enable color reproduction that is perceived by humans without darkening the impression of the image. And

An image processing apparatus according to an aspect of the present invention includes:
For each of the sections obtained by dividing the color space represented by hue, saturation, and lightness with respect to hue, the outline and saturation of the range in which pixel values can be distributed on the target hue plane represented by the hue value representing each section A maximum saturation point estimation unit for obtaining a maximum saturation point at which a possible value is maximized;
For each of the sections, a reference line setting unit that sets a reference line that increases in lightness as the saturation increases through the maximum saturation point and the lightness axis of the color space;
A correction direction calculation unit for obtaining a correction direction of each pixel value according to a distance between a point of the color space representing each pixel value of the input image and the reference line of the section to which each pixel belongs;
A correction unit that obtains a corrected pixel value by correcting each pixel value according to a distance from the point of the color space representing each pixel value of the input image to the outline in the correction direction;
An output unit for outputting the corrected pixel value;
Is provided.

An image processing method as one aspect of the present invention includes:
For each of the sections obtained by dividing the color space represented by hue, saturation, and lightness with respect to hue, the outline and saturation of the range in which pixel values can be distributed on the target hue plane represented by the hue value representing each section A maximum saturation point estimation step for obtaining a maximum saturation point at which a possible value is maximized;
For each of the sections, a reference line setting step for setting a reference line that increases in brightness as the saturation increases through the maximum saturation point and the brightness axis of the color space;
A correction direction calculating step for obtaining a correction direction of each pixel value according to a distance between a point of the color space representing each pixel value of the input image and the reference line of the section to which each pixel belongs;
A correction step of obtaining a corrected pixel value by correcting each pixel value according to a distance from the point in the color space representing each pixel value of the input image to the outline in the correction direction;
An output step of outputting the corrected pixel value;
Is provided.

  According to the present invention, it is possible to reproduce colors that humans feel preferable by effectively utilizing the color gamut.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First embodiment)
FIG. 1 is a block diagram schematically showing the configuration of the color image processing apparatus according to the first embodiment.

  The block diagram of FIG. 1 represents the units of each function obtained by causing a microprocessor to execute a program according to this embodiment. The program according to the present embodiment can be included as a module of an image processing module of application software for a computer including a display device such as a digital camera, a mobile phone, and a television. Here, the configuration in which the image processing device is realized by software is shown, but all or a part of the functions of the image processing device can also be configured by a hardware device.

  FIG. 2 is a flowchart showing a flow of processing performed by the image processing apparatus of FIG. The operation of each element of the image processing apparatus in FIG. 1 will be described below with reference to FIG.

  The input unit 101 in the image processing apparatus of FIG. 1 acquires image data from a storage device built in the computer (step S1). The acquired image data represents an M × N pixel image in which M pixels are arranged in the horizontal direction and N pixels are arranged in the vertical direction. This image is an RGB image, and each pixel has gradation data of red, green, and blue components. The gradation data of each component is expressed by 8 bits, for example. The data that can be processed is not limited to 8-bit RGB data, but may be any data that can be converted into a visually uniform space, such as sRGB data, YCC data, and XYZ data. Each channel is not limited to 8 bits.

  The color space conversion unit 102 converts the image data (that is, RGB space image data) acquired by the input unit 101 in step S1 into color space image data represented by three attributes (components) of brightness, saturation, and hue. (Step S2). The converted pixel value is referred to as a colorimetric value. The RGB space is an example of a first color space formed with a plurality of color components. A color space represented by three attributes of brightness, saturation, and hue corresponds to a second color space defined on a coordinate system represented by components of hue, saturation, and brightness.

  Although there are various spaces that can be represented by the three attributes of brightness, saturation, and hue, it is desirable that the spaces be visually uniform because operations such as obtaining the inclination in the color space are performed as described later. Here, a CIEELCH color space, which is a visually equivalent space and which expresses the CIELAB color space of the orthogonal coordinate system in a cylindrical coordinate format, is used. It can be said that the CIELCH color space is a color space in which it is possible to easily determine what kind of characteristic the color is. However, as the uniform color space to be processed, the present invention is not limited to the CIELAB color space, but may be an LUV color space, and HSV or the like is also used as long as it is a space that can be represented by three attributes of color, brightness, saturation, and hue. be able to.

  The vertex determination unit (maximum saturation point estimation unit) 103 is a target hue plane represented by a hue value representing each section for each of the sections obtained by dividing the color space represented by hue, saturation, and brightness with respect to hue. Thus, the outline of the range in which the pixel values can be distributed and the maximum saturation point (vertex) at which the value that the saturation can take are maximized (step S3). In this embodiment, a predetermined number of hue values arranged at predetermined intervals in the hue of the color space can be used, the predetermined number matches the number of divided sections, and the predetermined number of each hue value is a section. Is the same as the hue value representative of.

  Here, the outermost vertexes of each section (target hue plane) may be registered in the database in advance. Alternatively, the outermost vertexes may be discretely registered in the database for some sections (target hue planes). In this case, the outermost vertex for the unregistered hue may be determined by interpolation calculation.

  The reference inclination determination unit 104 determines, for each section, a reference line that passes through the outermost vertex and increases in brightness as the saturation increases (has a positive inclination) (step S4). The reference line is a line having an inclination (reference inclination) that serves as a reference when determining the correction direction of the colorimetric value described later.

  FIG. 3 is a schematic diagram for explaining processing for determining a reference line by paying attention to a certain section (target hue plane).

A saturation-lightness range corresponding to the one target hue plane is shown. The reference line is determined so as to pass through the outermost vertex (maximum saturation point) R and to have a positive slope. In the color space, the coordinates (three attributes) of the outermost vertex R are represented as (L _R , C _R , H _R ). L represents lightness, C represents saturation, and H represents hue. If the intersection point with the straight line extending perpendicularly to the brightness axis from the outermost vertex R is P, the intersection point P is (L _R , 0, H _R ). If the intersection of the reference line and the lightness axis is Q, the intersection Q is (L _Q , 0, H _R ). Therefore, the slope of the reference line can be expressed as (L _R −L _Q ) / C _R , and the intercept can be expressed as L _Q.

In general, when a color is widely distributed in a color gamut, a person tends to determine that the image is rich in gradation, and tends to prefer a moderately bright and vivid image. However, if it is too bright or too vivid, there is a tendency to judge that the color reproduction is worse. Even if the color gamut of the input image is expanded, the shape of the color gamut varies greatly depending on the hue, and the direction and amount of the color gamut varies depending on the three attributes (lightness, saturation, hue). For example, a bright color with a yellow hue has high brightness, but a bright color with a blue hue has low brightness. At the time of expansion, it is necessary to moderately increase the brightness in order to achieve preferable color reproduction along with expansion in the saturation direction. Note that the saturation can be expanded to the saturation of the outermost vertex (maximum saturation). Therefore, the reference line is always set to have a positive inclination so that the brightness increases as it goes from the brightness axis toward the direction in which the saturation increases in any hue. Since L _R > L _Q , (L _R −L _Q ) / C _R > 0.

Here, the method of determining the intercept L _Q, are various. Although the simplest method is to set an arbitrary value for each hue, for example, the following can be considered as another determination example.

First, an intercept L _Q0 for a reference hue (reference hue) is determined. The reference hue may be red, for example, with the outermost apex positioned approximately in the middle of the brightness width. Subjective evaluation was variously changed value of the intercept L _Q for previously prepared image (e.g. a gradation image), to determine the optimal value. Alternatively, the intercept L _Q0 may be determined by multiplying the lightness L _R0 of the outermost vertex in the reference hue by a coefficient. Other hues different from the reference hue are calculated by calculation using the intercept L _Q0 obtained for the reference hue. For example, the lightness L _R0 of the outermost vertex R ₀ (L _R0 , C _R0 , H _R0 ) in the reference hue and the lightness L of the outermost point R _i (L _Ri , C _Ri , H _Ri ) in the hue to be calculated. _Using the ratio to _Ri (L _Ri / L _R0 ) and the slope of the reference line corresponding to the reference hue ((L _R0 −L _Q0 ) / C _R0 ), L _Qi = ((L _R0 −L _Q0 ) / C _R0 ) × (L _Ri / L _R0 ).

Here, the lightness ratio (L _Ri / L _R0 ) is simply multiplied, but instead of multiplying the lightness ratio (L _Ri / L _R0 ), the lightness in the middle of the color space, that is, the CELELCH color space is intermediate. A difference between a value (for example, 50 when the maximum value is 100) and the brightness L _R0 may be multiplied by a ratio between the difference between the intermediate value and the brightness L _Ri . In this manner, the three-attributes of the vertices of the outermost of each color to calculate the intercept L _Q to reflect, in the conversion value calculator 106 described later, to perform color conversion in consideration of the respective colors of the color gamut shape Is possible.

Calculation Example of sections L _Q described above is an example, as another example, simply, for each hue, a coefficient times the brightness of the apex of the outermost may be sectioned L _Q.

  In the above description, a monotonically increasing straight line (linear function) is calculated as the reference line. However, the reference line is not limited to a straight line, and may be any other function as long as it is a function that monotonically increases at least between the brightness axis and the vertex.

  For each pixel of the input image, the inclination calculation unit 105 is the distance between the coordinate point representing the colorimetric value of each pixel and the reference line of the section to which each pixel belongs (this example assumes a distance in the lightness axis direction). Accordingly, the correction direction (conversion direction) of the colorimetric value of each pixel is calculated (step S5). In step S6, which will be described later, the colorimetric value is corrected based on the distance to the outermost contour in the correction direction determined in step S5.

  FIG. 4 is a schematic diagram illustrating processing for calculating a correction direction (conversion direction) of a colorimetric value of a certain pixel.

As the coordinate point of the colorimetric value moves away from the reference line in the brightness axis direction, the correction direction is calculated so that the inclination becomes smaller than the reference line (approaching parallel to the saturation axis), for example, The slope is constant (= 0) at the maximum brightness and the minimum brightness. The inclination in the correction direction is smoothly changed with the reference inclination as the maximum reference according to the distance of the coordinate point from the reference line. As a specific calculation method, for example, when the colorimetric values (L _i , C _i , H _i ) are located below the reference line, the lightness L _0Ci of the reference line in the chroma C _i of the colorimetric values. Ask for. Then, the inclination in the correction direction of the colorimetric _value is _obtained by multiplying the reference inclination by L _i / L _0Ci . Similarly, when the colorimetric values (L _i , C _i , H _i ) are located above the reference line, the reference inclination is multiplied by (L _max −L _i ) / (L _max −L _R ). The inclination of the colorimetric value correction direction is obtained. L _max is the maximum value of brightness. When the colorimetric value is located on the reference line, the inclination of the reference line at that position becomes the inclination in the correction direction. In addition, by using the obtained inclination, it is also possible to obtain an intercept (intersection with the lightness axis) of a straight line that passes the measured value and has the inclination.

  In FIG. 4, the slope does not become smaller than the fixed value (= 0). However, for example, when the lightness is small, that is, in the low lightness region, it is possible to make it smaller than 0. This is because, in the low lightness region, even if the lightness is somewhat dark, the preferred color reproduction is not greatly affected.

  The conversion value calculation unit 106 corrects the correction amount to be added to the colorimetric value based on the distance from the colorimetric value to the outermost contour in the correction direction obtained by the inclination calculation unit 105 in step S5 (because the hue is constant, Is calculated), and the correction amount is added to the colorimetric value, thereby obtaining a correction value of the colorimetric value. That is, the movement amount of the coordinate point of the colorimetric value is calculated in the correction direction, the coordinate point is moved by the calculated movement amount, and the value of the coordinate point after the movement becomes a correction value to be obtained. For example, the correction policy is as follows.

  In an area of very small saturation that is considered to be an achromatic color, when the color is widened in the color gamut (if the saturation increases), the sense of discomfort increases and the image quality deteriorates. Therefore, saturation should not be increased in a region where saturation is very small. On the other hand, the high saturation area should not protrude from the color gamut that can be displayed, and should not be crushed (not concentrated on the outermost contour). As for the lightness, it is desired that the color does not change so much due to the conversion in the high lightness or low lightness region. Each colorimetric value is corrected according to a calculation formula reflecting such a policy.

For example, the following items (1) to (4)
(1) Distance from the coordinates of colorimetric values to the outermost contour along the correction direction (conversion direction)
(2) Hue
(3) Lightness
(4) Saturation
Each of the membership functions is prepared, and the amount of movement (correction amount) in the correction direction is obtained by multiplying the output values of the membership functions. Then, the coordinate point of the colorimetric value is moved by the calculated amount of movement along the correction direction, and the value of the coordinate point after movement is obtained as the value after correction of the colorimetric value. This is performed for all colorimetric values. In this way, the movement amount (correction amount) can be determined fuzzyly from a plurality of membership functions. Although it has been described here that the correction amount is determined fuzzy, what is important is that the conversion is performed so as to be spatially continuous, which is an important condition for preventing the occurrence of tone jumps and artifacts. Here, the movement amount (correction amount) along the correction direction is acquired, but in addition to this, for example, the correction amount in the saturation direction is obtained from the membership functions (1) to (4), and The correction amount may be obtained by calculation from the colorimetric value, the correction amount in the saturation direction, and the inclination in the correction direction. Alternatively, the correction amount in the lightness direction is obtained from the membership functions (1) to (4), and the correction amount in the saturation direction includes the colorimetric value, the correction amount in the lightness direction, and the inclination in the correction direction. You may make it obtain | require by calculation from.

  The color space reverse conversion unit 107 reversely converts the corrected colorimetric value of each pixel output from the conversion value calculation unit 106 from the uniform color space to the original color space (here, RGB color space) ( Step S7). That is, the image data in the uniform color space is inversely converted to the original color space (here, the RGB color space).

  The output unit 108 outputs the image data after the inverse transformation obtained by the color space inverse transformation unit 107 (step S8). The output image data is displayed on a display panel (not shown), for example.

(Second Embodiment)
The present embodiment is characterized in that the reference line is determined using the color distribution information of the input image. The block diagram of the image processing apparatus according to the present embodiment is the same as that in FIG. 1, but the processing of the reference inclination determination unit 104 is extended.

  FIG. 5 is a flowchart showing a flow of processing performed by the image processing apparatus according to the present embodiment. In FIG. 5, compared with the flowchart of FIG. 2, step S10 is added between steps S2 and S3.

  First, as in the first embodiment, image data is acquired in step S1, and in step S2, color space data (colorimetric value data) represented by three attributes of lightness, saturation, and hue is used. Convert to

  In step S10 following step S2, the reference inclination determination unit 104 obtains the color distribution of the image using the converted image data. More specifically, the entire hue is divided into a plurality of sections with a predetermined hue width, and an average value (centroid value) or an intermediate value of each pixel is obtained as color distribution information for each section.

  In step S3 following step S10, for each section, the outline and the outermost vertex are determined on the target hue plane represented by the hue of the statistical value.

  In step S4, a reference line is determined so as to pass a point representing a statistical value (such as a centroid value or a center value) of each section and an outermost vertex (maximum again point).

  FIG. 6 is a diagram schematically showing processing for determining a reference line. Here, an example in which the center of gravity is used as the statistical value of the color distribution is shown, but other statistical values such as a central value may be used.

The center of gravity G obtained in step S10 for a certain section is defined as (L _G , C _G , H _G ). Also, let R be the vertex of the outermost contour in the hue of the center of gravity, and let its three attributes be (L _R , C _R , H _R ). At this time, a line passing through the outermost vertex R and the center of gravity G is determined as a common reference line for each pixel included in the section. By determining the reference line in this way, color conversion is performed using the color gamut effectively because the center of gravity is color-converted in the direction of the highest saturation.

However, the reference line connecting the outermost vertex and the center of gravity does not always have a positive slope (ie, L _G <L _R ). As described above, if the pixel value is converted to a darker direction, an image display that is preferable for human beings cannot be obtained. Therefore, when the L _G> L _R is, the process shown in the first embodiment, it is effective to determine the reference line.

  Steps S5 to S8 are the same as in the first embodiment. That is, in step S5, the inclination in the correction direction of the colorimetric value of each pixel is calculated. In step S6, the colorimetric value is corrected in the obtained correction direction. In step S7, the corrected colorimetric value of each pixel. Are converted back to the color space of the original image data, and the image data after the reverse conversion is output in step S8. Note that when the width of the hue range is large, the value of the inclination in the correction direction obtained in step S5 may be adjusted by interpolation calculation.

  As described above, according to the present embodiment, color conversion reflecting the content of image data is possible by determining the reference line using the color distribution characteristics of the image data. In the first embodiment, although color conversion is performed in consideration of the color gamut shape of each hue, the same color conversion is always performed for the same colorimetric value regardless of the contents of the input image data. . On the other hand, according to the present embodiment, even if the colorimetric value is the same, different color conversion is performed according to the content of the image data, and a more preferable image display for humans is possible.

(Third embodiment)
In the first embodiment, the image data is converted into data of a uniform color space, the colorimetric value of each pixel is acquired, the colorimetric value of each pixel is corrected, and the colorimetric value after each correction is converted to the original color. Inverted to space and output the image data after the inverse transformation. In this embodiment, a database in which colorimetric values of image data are associated with correction values obtained by correcting the colorimetric values is created in advance, and color conversion is performed using this database. Steps S3 to S6 are omitted, and the amount of calculation for color conversion is to be reduced.

  FIG. 7 is a block diagram schematically showing the configuration of the image processing apparatus (database creation apparatus) according to the present embodiment. Elements having the same names as those in FIG. 1 are denoted by the same reference numerals, and description thereof is omitted except for extended processing.

  FIG. 8 is a flowchart showing the flow of processing performed by the apparatus of FIG.

  In step S11, the input unit 101 acquires colorimetric values. As described above, the colorimetric value is a pixel value in a uniform color space.

  Steps S3 to S6 are the same as steps S3 to S6 of FIG. 2 used in the first embodiment, and corrected colorimetric values are obtained for each pixel by these steps.

  In step S18, the output unit 108 creates the database 110 by associating the colorimetric values before correction with the colorimetric values after correction, and stores the created database 110 in a storage device (not shown).

  When processing actual image data, the elements 103 to 106 in FIG. 1 are replaced by the database search 111. FIG. 9 is a block diagram, and FIG. 10 is a flowchart. Elements having the same names as those in FIGS. 1 and 2 are denoted by the same reference numerals, and description thereof will be omitted except for expanded processing.

  In step S21, the database retrieval unit 111 retrieves each pixel value that has undergone color space conversion of the input image data, and outputs the corrected pixel value. Here, the database 110 can be described in a one-to-one correspondence between input and output, but in that case, the database has a huge capacity. Therefore, only a few points (pixel values) may be described, and the database scale may be reduced by calculating the points between them by interpolation calculation. Thereby, it is possible to display an image more preferable for humans while reducing the amount of calculation.

1 is a diagram showing a configuration of an image processing apparatus according to a first embodiment of the present invention. 2 is a flowchart showing a flow of processing by the image processing apparatus in FIG. 1. The schematic diagram explaining the process of a reference | standard inclination determination part. The schematic diagram explaining the process of an inclination calculation part. 9 is a flowchart showing a flow of processing by an image processing apparatus according to a second embodiment of the present invention. The schematic diagram explaining the process of the reference | standard inclination determination part in 2nd Embodiment. The figure which shows the structure of the image processing apparatus (database creation apparatus) which concerns on 3rd Embodiment of this invention. 8 is a flowchart showing a flow of processing by the image processing apparatus of FIG. The figure which shows the structure of the image processing apparatus which concerns on 3rd Embodiment of this invention. 10 is a flowchart showing a flow of processing by the image processing apparatus of FIG. The schematic diagram which expands the saturation of a pixel in connection with a prior art. The schematic diagram which expands a pixel value centering | focusing on the intermediate point of a brightness axis in connection with a prior art. The a * b * chromaticity diagram when the sRGB color space is represented by the CIELAB color space of the uniform color space. Color gamut shape of each hue of yellow, blue and red in the sRGB color space. The schematic diagram which expands a pixel value to the vertex direction.

Explanation of symbols

101: input unit 102: color space conversion unit 103: vertex determination unit 104: reference inclination determination unit 105: inclination calculation unit 106: conversion value calculation unit 107: color space inverse conversion unit 108: output unit

Claims (8)

For each of the sections obtained by dividing the color space represented by hue, saturation, and lightness with respect to hue, the outline and saturation of the range in which pixel values can be distributed on the target hue plane represented by the hue value representing each section A maximum saturation point estimation unit for obtaining a maximum saturation point at which a possible value is maximized;
For each of the sections, a reference line setting unit that sets a reference line that increases in lightness as the saturation increases through the maximum saturation point and the lightness axis of the color space;
A correction direction calculation unit for obtaining a correction direction of each pixel value according to a distance between a point of the color space representing each pixel value of the input image and the reference line of the section to which each pixel belongs;
A correction unit that obtains a corrected pixel value by correcting each pixel value according to a distance from the point of the color space representing each pixel value of the input image to the outline in the correction direction;
An output unit for outputting the corrected pixel value;
An image processing apparatus comprising: The correction direction calculation unit calculates the correction direction so that the inclination becomes smaller than the reference line as the distance to the reference line increases.
The image processing apparatus according to claim 1. The reference line setting unit obtains the reference line passing through the center of gravity of the pixel values belonging to each section in the color space and the maximum saturation point,
The image processing apparatus according to claim 1, wherein the center of gravity of the pixel value belonging to each section corresponds to a hue value representing each section. The reference line setting unit obtains the reference line passing through the point of the color space corresponding to the intermediate value of the pixel values belonging to each section and the maximum saturation point;
The image processing apparatus according to claim 1, wherein the intermediate value of the pixel values belonging to each section corresponds to a hue value representing each section. For each of the sections obtained by dividing the color space represented by hue, saturation, and lightness with respect to hue, the outline and saturation of the range in which pixel values can be distributed on the target hue plane represented by the hue value representing each section A maximum saturation point estimation step for obtaining a maximum saturation point at which a possible value is maximized;
For each of the sections, a reference line setting step for setting a reference line that increases in brightness as the saturation increases through the maximum saturation point and the brightness axis of the color space;
A correction direction calculating step for obtaining a correction direction of each pixel value according to a distance between a point of the color space representing each pixel value of the input image and the reference line of the section to which each pixel belongs;
A correction step of obtaining a corrected pixel value by correcting each pixel value according to a distance from the point in the color space representing each pixel value of the input image to the outline in the correction direction;
An output step of outputting the corrected pixel value;
An image processing method comprising: The correction direction calculation step calculates the correction direction so that the inclination becomes smaller than the reference line as the distance to the reference line increases.
The image processing method according to claim 5. The reference line setting step obtains the reference line passing through the center of gravity of the pixel value belonging to each section in the color space and the maximum saturation point,
The image processing method according to claim 5, wherein the center of gravity of the pixel value belonging to each section corresponds to a hue value representing each section. The reference line setting step obtains the reference line passing through the point of the color space corresponding to the intermediate value of the pixel values belonging to each section and the maximum saturation point,
The image processing method according to claim 5, wherein the intermediate value of the pixel values belonging to each section corresponds to a hue value representing each section.

Images