JP2008148163A - Color processing apparatus and method thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To obtain unified color reproduction between a plurality of similar images. <P>SOLUTION: An image processing apparatus performs color processing for converting image data represented in a first color gamut into image data of a second color gamut different from the first color gamut. In that case, an image input unit 106 registers a plurality of image data items in an image database 112 by categories. An image feature amount calculation unit 107 calculates an image feature amount of image data, for each category, registered in the image database 112. In response to a user instruction based on the image feature amount, a mapping method selection unit 108 sets color processing to be applied to the image data by categories. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to color processing for converting image data represented by a first color gamut into image data of a second color gamut different from the first color gamut.

  A color management system (Color Management System: CMS) is known as a color management technique for satisfactorily reproducing one color image in a plurality of image input / output devices. The CMS uses a source profile (predetermined conversion formula or conversion table) that defines color conversion from a color in a device-dependent color space to a color in a device-independent color space. Further, a destination profile that defines color conversion from a color in a device-independent color space to a color in a device-dependent color space is used. Color management is realized by these two color conversions. For example, CIELab or CIEJCh space is used as a device-independent color space.

  However, when the output color gamut (hereinafter referred to as the output color gamut) is narrower than the input color gamut (hereinafter referred to as the input color gamut), it is impossible to accurately reproduce the input color information. There are cases. For example, when an image on a monitor with a wide color gamut is output by a printer with a narrow color gamut, colors outside the color gamut of the printer cannot be reproduced as they are. In such a case, it is necessary to perform color processing that satisfactorily reproduces the color outside the color gamut with the color within the color gamut while maintaining the original image information (gradation, hue, etc.) as much as possible. There is color gamut mapping as a technique for realizing this.

  Various methods have been proposed for color gamut mapping. 1A to 1D are diagrams illustrating a typical process for performing color gamut mapping on a two-dimensional plane of lightness and saturation while fixing a hue. FIG. 1A shows a color gamut mapping method in which the lightness is fixed and the saturation is compressed. FIG. 1B shows a color gamut mapping method that compresses lightness with constant saturation. FIG. 1C shows a mapping method that minimizes the amount of change before and after mapping lightness and saturation. FIG. 1D shows a mapping method for compressing lightness and saturation toward a predetermined fixed point (focal point) P on the lightness axis.

  In addition to the mapping methods shown in FIGS. 1A to 1D, there is known a method of reducing the compression ratio (change amount before and after mapping) of lightness and saturation by changing the hue within a predetermined range. In order to maintain gradation, the color signal value is maintained up to a predetermined non-compressed area in the color gamut, and the color signal value exceeding the non-compressed area is determined by using a predetermined compression function. There is known a method of smoothly mapping between the boundary surface of the output color gamut and the output color gamut.

  However, each of these mapping methods has advantages and disadvantages. For example, when color gamut mapping is performed with a constant hue for an image including many colors outside the color gamut, the saturation is preserved, so that increasing the compression of lightness lowers the contrast and loses the stereoscopic effect of the image. On the contrary, in order to preserve the lightness, if the compression of the saturation is increased, the vividness is lowered and the image has no impact.

  In addition, the mapping method for maintaining the color signal values up to the non-compressed region is independent of the distribution of the color signal values of the input image. Although it can be reproduced faithfully, it may be mapped.

  Thus, it is very difficult to obtain a good color reproduction with one mapping method for all images.

  As an image adaptive mapping method, there is known a method of calculating a feature amount of an image to be mapped and performing optimum color gamut mapping based on the feature amount. However, in order to optimize the color gamut mapping parameters according to the shooting scene of the images, even if multiple similar images have the same subject color data, they are treated as different subject color data among the images. Therefore, there is a problem that uniform color reproduction cannot be obtained.

JP-A-7-107306

  An object of the present invention is to obtain a uniform color reproduction between a plurality of similar images.

  The present invention has the following configuration as one means for achieving the above object.

  When performing color processing to convert image data represented by the first color gamut into image data of the second color gamut different from the first color gamut, a plurality of image data is registered in the memory by category. Then, an image feature amount of the image data for each category registered in the memory is calculated, and color processing to be performed on the image data is set for each category based on the image feature amount.

  According to the present invention, uniform color reproduction can be obtained between a plurality of similar images.

  Hereinafter, image processing according to an embodiment of the present invention will be described in detail with reference to the drawings.

[Device configuration]
FIG. 2 is a block diagram illustrating a configuration example of the image processing apparatus 101 according to the first embodiment.

  In FIG. 2, an image display device 102 is a monitor or the like for displaying an image. The image output device 103 is a printer or the like for outputting an image. The image input device 104 is a digital camera or the like for inputting an image. Each of these apparatuses is connected to the image processing apparatus 101 via a predetermined interface (for example, USB or IEEE1394).

  In the image processing apparatus 101, the image display unit 105 generates a video signal to be displayed on the image display apparatus 102. The image input unit 106 stores the digital image acquired by the image input device 104 in the image database 112. The image output unit 111 generates an image signal for outputting an image by the image output device 103.

  The image feature amount calculation unit 107 calculates the feature amount of the image registered in the image database 112. The mapping method selection unit 108 selects a color gamut mapping method suitable for each image registered in the image database 112. The color processing unit 109 performs color processing for outputting an image by the image output device 103.

  The input profile storage unit 113 stores source profiles of the image display device 102, the image input device 104, and the like. Destination profiles such as the output profile storage unit 114 and the image output device 103 are stored. Note that the image database 112, the input profile storage unit 113, and the output profile storage unit 114 are allocated to a nonvolatile memory such as a hard disk drive, for example. The data buffer 115 is a RAM or the like that temporarily stores data for data processing.

  A user interface (UI) unit 116 generates image data for displaying a user interface for the user to operate the image processing apparatus 101 on the image display apparatus 102. The operation input unit 110 is, for example, a keyboard, a pointing device, a touch panel, or the like for inputting user instructions such as color data targeted for a specific color.

[Image processing]
FIG. 3 is a flowchart for describing image processing executed by the image processing apparatus 101.

  The image processing apparatus 101 first performs an initialization process such as initializing the data buffer 115 (S201).

  Next, the image processing apparatus 101 inputs the image data acquired by the image input apparatus 104 via the image input unit 106, classifies the image data according to the category specified by the user, and registers the image data in the image database 112 (S202). .

  Next, the image processing apparatus 101 reads the image data of the category specified by the user registered in the image database 112 into the data buffer 115 and calculates the image feature amount (color distribution characteristic) of the image data to the image feature amount calculation unit 107. (S203). A method for calculating the image feature amount by the image feature amount calculation unit 107 will be described later.

  Next, the image processing apparatus 101 acquires color reproduction information of the image input apparatus 104 and the image output apparatus 103 from the input profile storage unit 113 and the output profile storage unit 114 in accordance with a user instruction (S204). The color reproduction information acquired here describes the conversion characteristics from the device-dependent color space (eg, RGB space) included in the source profile to the device-independent color space (eg, XYZ, Lab color space, etc.). It is a data table. Furthermore, the data table describes conversion characteristics from a device-independent color space included in the destination profile to a device-dependent color space (for example, CMYK space).

  The data table of the first embodiment describes the correspondence between the RGB values of the grid points and the corresponding Lab values as in the format shown in FIG. 5, for example. FIG. 5 shows an example in which the combination of RGB values and Lab values of 9 × 9 × 9 grid points arranged uniformly is described. Of course, the arrangement and number of grid points are not limited to 9 × 9 × 9, but may be non-uniform, 5 × 5 × 5, 7 × 7 × 7, or the like.

  Next, the image processing apparatus 101 reads target color data in accordance with a user instruction to be described later (S205). FIG. 6 is a diagram illustrating an example of target color data. The target color data describes the correspondence between RGB values and Lab values indicating the colors (important colors) that the user wants to reproduce.

  Next, the image processing apparatus 101 inputs a user instruction (to be described later) from the mapping method selection unit 108, and creates a mapping method table describing the relationship between the name of the image data (for example, the data file name), the category, and the mapping method. Update (S206). Although details will be described later, the user selects an optimal mapping method with reference to the color distribution characteristics of image data by category.

  FIG. 7 shows an example of the mapping method table. The mapping method table is stored in a predetermined area of the image database 112.

  Next, the image processing apparatus 101 converts the image data instructed by the user from the RGB value to the Lab value by using the color reproduction information of the image input apparatus 104 acquired in step S204 by the color processing unit 109 (S207). . Then, the color processing unit 109 uses the relationship between the color gamut of the image input device 104 and the color gamut of the image output device 103 indicated by the color reproduction information acquired in step S204, and the mapping method selected in step S206 to generate image data. The Lab values are color gamut mapped (S208). Then, using the color reproduction information of the image output device 103 acquired in step S204, the color processing unit 109 converts the Lab data from the color gamut mapped image data into the CMYK value of the image output device 103 (S209).

  In steps S207 and S208, based on the target color data read in step S205, target color data of a specific color important for the user is set, and color processing is performed to output the target color data. In steps S207 and S208, interpolation calculation such as tetrahedron interpolation is used for calculation of signal values other than the grid points having no table data.

  Next, the image processing apparatus 101 causes the image output unit 111 to generate print data that can be processed by the image output apparatus 103 from the image data converted in step S208 and output the print data to the image output apparatus 103 (S209).

  In FIG. 3, the process from the image data registration process (S202) to the image data output process (S210) is described as a series of processes. However, in practice, the image data registration process (S202), the feature value calculation process (S203), the mapping method table update process (S206), and the image data color process (S207) to the image data output process (S210). Can be divided. These processes are executed independently by operation of a user interface described later.

[User interface]
FIG. 4 is a diagram illustrating an example of a UI for operating image processing, which is displayed on the image display device 102 by the image processing device 101.

  In the area 301, a radio button 302 for the user to select a category for calculating the image feature amount of the image data is arranged. The user can set an arbitrary category name for the image data when registering the image data in the image database 102. That is, the category names set in the image data registered in the image database 102 are listed in the area 301.

  In the area 303, a histogram indicating the image feature amount (color distribution characteristic) of the image data of the category designated by the user is displayed. By selecting the button 310, the user can display each of the lightness distribution, the saturation distribution, and the hue distribution, or a histogram in which two or three are overlaid. In addition, the user can select color regions such as R, G, B, C, M, and Y using the button 311 and display the brightness and saturation histograms of each region.

  The image feature amount calculation button 304 is a button for instructing the calculation of the image feature amount of the image data of the category designated by the user.

  A drop-down menu 305 is a menu for the user to select a mapping method from mapping methods prepared in advance. For example, when printing a photographic image of the sea, the user refers to the lightness distribution, the saturation distribution, and the hue distribution of the image data classified by the sea category. If the color distribution is biased in the cyan area, the mapping method B (cyan area storage) is selected, and the mapping method that places importance on the color reproducibility of the cyan area is selected.

  FIG. 7 is a diagram showing a table describing the relationship between image data names (for example, data file names), categories, and mapping methods, which is created by the user's mapping method selection. This table is stored in a predetermined area of the image database 112.

  A print image data designation button 306 is a button for designating image data that the user desires to print.

  The color reproduction information acquisition button 307 is a button for the user to specify names of the image input device 104 and the image output device 103. In step S204, the color reproduction information of the image input device 104 and the image output device 103 corresponding to the name specified by the operation of the button 307 is acquired. Note that the name and the like of the image input device 104 can also be acquired from information added to the image data. In this case, the user does not need to operate the color reproduction information acquisition button 307 and specify the name or the like of the image input device 104.

  The target color data acquisition button 308 is a button for instructing acquisition of data in the format shown in FIG. 6 in which the user describes important colors.

  The image output button 309 is a button for instructing the user to print image data. When this button 309 is pressed, the image data designated by operating the print image data designation button 306 is subjected to color processing including color gamut mapping by the mapping method described in the table shown in FIG. The data is output to the output device 103.

[Image feature calculation]
FIG. 8 is a flowchart for explaining the processing of the image feature amount calculation unit 107.

  The image feature quantity calculation unit 107 performs, as preprocessing, sampling of pixels of image data and conversion of an unspecified image size to a predetermined size in order to reduce the calculation cost due to histogram creation (S701).

  Next, the image feature amount calculation unit 107 reads the source profile of the image input device 104 from the input profile storage unit 113 in accordance with the name specified by the operation of the color reproduction information acquisition button 307 or information added to the image data. (S702).

Next, the image feature amount calculation unit 107 selects one of the image data stored in the data buffer 115 (S703). Then, one pixel of the selected image data is extracted (S704), the RGB value of the pixel is converted into an Lab value using the source profile read in step S702, and the brightness L, color saturation is further expressed using equation (1). The degree C and the hue H are calculated (S705).
L = L *
C = √ (a * ² + b * ² )… (1)
H = tan ^-1 (b * / a *)

Next, the image feature amount calculation unit 107 updates the histogram stored in the predetermined area of the data buffer 115 with the lightness L, saturation C, and hue H calculated in step S702 (S706). The histogram includes a lightness L histogram HistL, a saturation C histogram HistC, and a hue H histogram HistH, and all elements of the histogram in the initial state are 0. The histogram is updated according to the equation (2).
HistL [L] = HistL [L] + 1
HistC [C] = HistC [C] + 1… (2)
HistH [H] = HistH [H] + 1

  Next, the image feature amount calculation unit 107 determines whether or not the creation of the histogram of all the pixels of one image data has been completed (S707). Then, the processes of steps S704 to S706 are repeated until the creation of the histogram of all the pixels of one image data is completed.

  When the creation of the histogram of all the pixels of one image data is completed, it is determined whether or not the creation of the histogram of all the image data stored in the data buffer 115 (all the image data of the specified category) is finished (S708). . Then, the processes from step S703 to S707 are repeated until the creation of the histogram of all image data of the specified category is completed.

  FIG. 9 is a diagram illustrating an example of a histogram created by the image feature amount calculation unit 107. In FIG.

  In this way, a plurality of image data can be made into a database for each category, color distribution characteristics of the plurality of image data can be analyzed, and color processing according to the user's purpose can be performed. Therefore, uniform color reproduction can be obtained between images having the same category.

  The image processing according to the second embodiment of the present invention will be described below. Note that the same reference numerals in the second embodiment denote the same parts as in the first embodiment, and a detailed description thereof will be omitted.

  In the first embodiment, the user refers to the color distribution characteristics displayed in the area 303 and uses the mapping method.

  The example of selecting an appropriate color gamut mapping method using the selection menu 305 has been described. Hereinafter, a process in which the mapping method selection unit 108, not the user, selects an appropriate color gamut mapping method will be described.

  In the following, as a color gamut mapping method, an uncompressed area in which the color gamut is not compressed is provided for each of the R, G, B, C, M, and Y areas. The color signal value is maintained up to the non-compressed region, and the color signal value exceeding the non-compressed region is smoothed between the boundary surface of the non-compressed region and the output color gamut using a predetermined compression function. A method of mapping will be described.

  FIG. 10 shows an uncompressed area based on the image feature amount after the image feature amount of the image data (image data of the specified category) stored in the data buffer 115, which is executed by the mapping method selection unit 108, is calculated. It is a flowchart explaining the method to set.

  The mapping method selection unit 108 first performs an initialization process (S901).

  Next, the mapping method selection unit 108 acquires the color reproduction information of the image output device 103 from the output profile storage unit 114 in accordance with the name specified by operating the color reproduction information acquisition button 307 (S902).

  Next, the mapping method selection unit 108 selects one of the hue regions of R, G, B, C, M, and Y from the color distribution information of the image data (image data of the specified category) stored in the data buffer 115. For this, the color signal value Cmax having the maximum saturation value is acquired (S903). It is assumed that the color distribution information has been calculated by the processing in the first embodiment.

  Next, the mapping method selection unit 108 initializes the control value N of the histogram to 0 (S904), and determines whether Cmax is within the output color gamut of the image output device 103 (S905).

  When Cmax is out of the output color gamut, the mapping method selection unit 108 increments the control value N (S906). Then, the color signal value having the saturation value having the highest appearance frequency in the upper N% of the histogram of the hue region for which Cmax is acquired in step S903 is set to Cmax (S907), and the process returns to step S905. Accordingly, steps S905 to S907 are repeated until Cmax is within the output color gamut.

  When Cmax falls within the output color gamut, mapping method selection section 108 sets the current Cmax to the uncompressed area of the hue area from which Cmax was acquired in step S903 (S908). For setting the uncompressed area, a method of setting the uncompressed area relative to the output color gamut of the image output apparatus 103 is used. That is, the color signal value having the maximum saturation is searched for in the color signal value of the image data (image data of the specified category) stored in the data buffer 115 and the common color gamut of the output color gamut of the image output device 103. To do. Then, the lightness value of the color signal value is obtained, and an uncompressed area of saturation is set on the lightness value.

  Next, the mapping method selection unit 108 determines whether or not the processing of steps S903 to S908 has been performed for all hue regions of R, G, B, C, M, and Y (S909). The processes in steps S903 to S908 are repeated.

[Modification]
In the above-described embodiment, an example has been described in which a histogram of brightness, saturation, and hue is created as the image feature amount of image data by category. However, the color distribution characteristics of the ab plane in the CIELab space may be used. Alternatively, the Lab signal value may be displayed in a pseudo three-dimensional manner, and the color gamut mapping method may be selected by referring to the three-dimensional color distribution characteristics.

  In the above embodiment, a color correction lookup table (color correction LUT) is generated using a user-selected color gamut mapping method, and color processing is performed using the color correction LUT. However, the color correction LUT corresponding to the image output device 103 and the output setting is stored in the nonvolatile memory in advance for each category, and the color correction LUT stored in the nonvolatile memory is referred to when performing color processing. May be.

  In the above embodiment, the mapping color space is Lab space, but for other uniform color space such as CIEJCH space, non-uniform color space such as XYZ space, or device-dependent RGB color space. However, the present invention is applicable.

  .so canon _3

A diagram showing a typical process for performing color gamut mapping on a two-dimensional plane of lightness and saturation with a fixed hue, A diagram showing a typical process for performing color gamut mapping on a two-dimensional plane of lightness and saturation with a fixed hue, A diagram showing a typical process for performing color gamut mapping on a two-dimensional plane of lightness and saturation with a fixed hue, A diagram showing a typical process for performing color gamut mapping on a two-dimensional plane of lightness and saturation with a fixed hue, Block diagram showing a configuration example of the image processing apparatus of Example 1, A flowchart illustrating image processing executed by the image processing apparatus; The figure which shows an example of UI for operating image processing, A diagram explaining color reproduction information, The figure which shows an example of target color data, The figure which shows an example of a mapping method table, A flowchart for explaining processing of the image feature amount calculation unit; The figure which shows the example of the histogram which an image feature-value calculation part produces, 10 is a flowchart illustrating a method for setting an uncompressed area, which is executed by a mapping method selection unit according to the second embodiment.

Claims (8)

A color processing device that converts image data represented by a first color gamut into image data of a second color gamut different from the first color gamut,
Registration means for registering a plurality of image data in a memory by category;
Calculation means for calculating the image feature amount of the image data by category registered in the memory;
A color processing apparatus comprising: setting means for setting color processing to be performed on image data for each category based on the image feature amount.   2. The color processing apparatus according to claim 1, wherein the calculating unit calculates a color distribution characteristic of the image data for each category as the image feature amount.   3. The color processing apparatus according to claim 2, wherein the color distribution characteristics include brightness, saturation, and hue distribution characteristics.   2. The color processing apparatus according to claim 1, further comprising selection means for selecting a color gamut mapping method in the color processing based on the image feature amount. A selecting means for selecting image data to be subjected to the color processing from the image data registered in the memory;
2. The color processing apparatus according to claim 1, further comprising color processing means for performing color processing corresponding to the category of the selected image data on the image data. A color processing method for converting image data represented by a first color gamut into image data of a second color gamut different from the first color gamut,
Register multiple image data in memory by category,
Calculating an image feature amount of the image data for each category registered in the memory;
A color processing method comprising: setting color processing to be performed on image data for each category based on the image feature amount.   7. A computer program for controlling the image processing apparatus to execute the color processing according to claim 6.   8. A computer-readable storage medium in which the computer program according to claim 7 is recorded.

Images