JP2004120092A - Image processing apparatus, image processing system, image processing method, storage medium, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus capable of easily and efficiently providing an image after excellent color processing even when an image of a processing object is any type of image. <P>SOLUTION: When the object image is divided into regions and color processing is applied to the division regions, division means 407 to 409 divide the object image into the regions on the basis of color information of the object image and spatial frequency information obtained by edge extraction applied to the object image. <P>COPYRIGHT: (C)2004,JPO

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention is, for example, an image processing apparatus, an image processing system, which is used for an apparatus or a system that divides an image and performs color processing such as color adjustment or color correction for each area or a necessary area. The present invention relates to an image processing method, a computer-readable storage medium storing a program for performing the image processing method, and the program.
[0002]
[Prior art]
Conventionally, for example, image processing of dividing an image into regions, and performing color processing such as color adjustment or color correction for each region or a necessary region is performed by creating a catalog or a flyer in a printing plate making process, This process is frequently performed on a designer side such as DTP (Desk Top Publishing) by image adjustment using image retouching software.
[0003]
By the way, in image processing in which an image is divided so that similar colors become the same area, and then color processing is performed on each area, an image to be processed is drawn on the background with characters that can be distinguished from the background color. In many cases, there is no problem if the image is easy to be divided into regions, such as a divided image. However, if the image to be processed is an image such as a natural image, specifically, if the image has a gradual color change, even if the image is divided into regions and color processing is performed, the region is divided. In some cases, the color processing performed on the divided area may cause a problem because the result does not match the human sense.
[0004]
As a typical example of the above-mentioned inconvenience, for example, as a result of performing color processing for each divided region, a dividing line dividing the region is conspicuous, and is a so-called boundary line that did not exist in the original image. There is a problem that a pseudo contour occurs and an unnatural processed image is obtained even when compared with the original image.
[0005]
Therefore, in order to solve the above-mentioned problems, as an image processing method aiming to increase the accuracy of region division and perform region division close to human sensation, for example, JP-A-6-237372 and the like A method has been proposed in which a space for performing clustering is devised so that the result of region division is closer to human senses.
[0006]
In addition, for example, in Japanese Patent Application Laid-Open No. H11-306334, a target image is clustered on a color and coordinate space to divide an area, and an area where a color change between the divided areas is small is set as an ambiguous area. Regarding the ambiguous area, a method of adjusting image processing parameters has been proposed.
[0007]
[Problems to be solved by the invention]
However, in the conventional image processing method described in Japanese Patent Application Laid-Open No. 6-237372 and the like, it seems that it is quite difficult to perform a region division close to a human sense, no matter how the region division process is devised. .
Further, in the conventional image processing method described in Japanese Patent Application Laid-Open No. H11-306334, it is necessary to perform color processing on the entire original image in order to perform region division processing with high accuracy. Both the required memory capacity and the amount of calculation are enormous. For example, in a case where an image is divided into regions and different image processing is performed for each of the divided regions, pre-scanning and region division at an actual resolution of the image and image processing (real resolution) for each image region are performed twice. Scanning at the actual resolution is required.
[0008]
Therefore, the present invention has been made to eliminate the above-mentioned drawbacks, and is capable of easily and efficiently providing a good color-processed image regardless of the image to be processed. It is an object to provide an apparatus, an image processing system, an image processing method, a computer-readable storage medium storing a program for executing the image processing method, and the program.
[0009]
Specifically, for example, even if an image to be processed is divided into regions, and then color processing is performed for each region, a pseudo contour is not generated at a region switching portion. In addition, when performing area division, the amount of memory and the amount of calculation required for this are reduced, so that simple area division can be performed.
[0010]
[Means for Solving the Problems]
For this purpose, the present invention relates to an image processing apparatus that divides a target image into regions and performs color processing on the divided regions, and performs color processing on the target image based on color information and spatial frequency information obtained from the target image. It is characterized by comprising a dividing means for dividing an image into regions.
[0011]
The present invention is also an image processing apparatus for performing a correction process on a target image based on characteristics of the target image detected by pre-scanning the target image, wherein the target image input at the time of the pre-scan is provided. For the areas that are not determined by the area division, a dividing unit that performs the area division again at the time of the above-described correction processing, and a division area obtained by the above-described division unit, according to the target area. Processing means for performing the correction processing.
[0012]
Further, the present invention is an image processing method for performing area division so that similar colors become the same area in a target image, comprising: an input step of inputting a target image; and a target image input by the input step. A color information recognizing step of recognizing the color information of the target image; a spatial frequency information recognizing step of recognizing the spatial frequency information of the target image input by the inputting step; a color information recognized by the color information recognizing step; An image area recognizing step of recognizing a target image area based on the spatial frequency information recognized in the information recognizing step.
[0013]
Further, the present invention is an image processing method for pre-scanning a target image to detect characteristics of the target image, and performing a correction process on the target image based on the detection result. And performing a region division again on the region not determined by the region division at the time of the pre-scan in the correction process, and performing the correction process according to the divided region. It is characterized by.
[0014]
Further, the present invention is a program for causing a computer to function as predetermined means, wherein the predetermined means divides a target image into regions and performs color processing on the divided regions. The present invention is characterized in that a means is provided for dividing a target image into regions based on color information and spatial frequency information obtained from the target image.
[0015]
Further, the present invention is a program for causing a computer to function as predetermined means, wherein the predetermined means corrects a target image based on characteristics of the target image detected by pre-scanning the target image. As means provided in the apparatus or system that performs the processing, the target image input at the time of the pre-scan is divided into regions, and the region not determined by the region division is divided again during the correction process. And a processing unit that performs the correction process according to the target area for each of the divided areas obtained by the dividing means.
[0016]
The present invention is also a program for causing a computer to execute a predetermined step, wherein the predetermined step is a processing step for dividing a target image into regions having similar colors. An input step of inputting, a color information recognition step of recognizing color information of the target image input by the input step, and a spatial frequency information recognition step of recognizing spatial frequency information of the target image input by the input step. And an image area recognition step of recognizing a target image area based on the color information recognized in the color information recognition step and the spatial frequency information recognized in the spatial frequency information recognition step.
[0017]
Further, the present invention is a program for causing a computer to execute a predetermined step, wherein the predetermined step detects a characteristic of the target image by pre-scanning the target image, and detects the characteristic of the target image based on the detection result. As a processing step for performing the correction processing, the area division of the input image is performed at the time of the pre-scan, and the area not determined by the area division at the time of the pre-scan is divided again at the time of the correction processing. And a processing step of performing the correction process according to the divided area.
[0018]
Further, the present invention is characterized in that the program according to any one of claims 14 to 17 is recorded on a computer-readable storage medium.
[0019]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0020]
The present invention is applied to, for example, a host computer 100 as shown in FIG.
The host computer 100 according to the present embodiment performs color processing such as color adjustment and color correction on an area obtained by dividing an image so that similar colors are the same, or performs color adjustment and color correction on each area obtained by dividing an image. It has a function of performing color processing such as correction, and in particular, even if color processing is performed for each area after the image is divided into areas, a pseudo contour does not occur at the switching part of the area, and a memory required for the area division It is configured to perform simple area division with a small amount and a small amount of calculation.
Hereinafter, the configuration and operation of the host computer 100 according to the present embodiment will be specifically described.
[0021]
<Configuration of Host Computer 100>
As shown in FIG. 1, for example, a printer 106 such as an inkjet printer and a monitor 105 are connected to the host computer 100.
[0022]
Further, the host computer 100 includes application software 101 such as a word processor, a spreadsheet, or an Internet browser, an OS (Operating System) 102, and various renderings indicating output images issued to the OS 102 by the application software 101. A printer driver 103 that processes commands (image drawing command, text drawing command, graphics drawing command, etc.) to create print data, and a monitor 106 that processes various drawing commands issued by the application software 101 And a monitor driver 104 for displaying a screen as software.
[0023]
The host computer 100 includes a central processing unit (CPU) 108, a hard disk driver (HD) 107, a random access memory (RAM) 109, and a read-only memory (ROM) as hardware on which the software 101 to 104 can operate. ) 110 and the like.
[0024]
In the present embodiment shown in FIG. 1, for example, a commonly used IBM compatible AT compatible personal computer is used as the host computer 100, and this personal computer is used in conjunction with Microsoft Windows. (R) 98 is used as the OS 102, an arbitrary printable application 101 is installed, and the monitor 105 and the printer 106 are connected.
[0025]
<A series of operations of the host computer 100>
In the host computer 100, for example, the CPU 108 starts the application 101 in response to an instruction from a user (an instruction based on an image displayed on the monitor 105). As a result, the application 101 can output image data using text data classified as text such as characters, graphics data classified as graphics such as graphics, or image image data classified as natural images or the like. Generate
[0026]
When printing output image data, the application 101 issues a print output request to the OS 102.
Specifically, for example, regarding the output image data, the application 101 sends, to the OS 102, a drawing command group indicating an output image including a graphics drawing command for the graphics data portion and an image drawing command for the image data portion. Issue.
[0027]
The OS 102 receives a drawing command group from the application 101, and issues the drawing command group as a print request to the printer driver 103 corresponding to the printer 106.
[0028]
The printer driver 103 generates print data printable by the printer 105 by processing a print request (drawing command group) from the OS 102, and transfers the print data to the printer 106.
More specifically, for example, when the printer 106 is a raster printer, the printer driver 103 sequentially performs image correction processing on a group of drawing commands from the OS 102, sequentially rasterizes the drawing command group into a 24-bit RGB page memory, and stores all drawing commands. Is converted into a data format printable by the printer 106, for example, CMYK data, and transferred to the printer 106.
[0029]
<Configuration of Printer Driver 103>
The printer driver 103 executes, for example, color processing as shown in FIG. 22, and includes an image correction processing unit 201 and a printer correction processing unit 202 as shown in FIG.
[0030]
The image correction processing unit 201 performs an image correction process on the color information included in the rendering command group from the OS 102.
Specifically, for example, the image correction processing unit 201 divides the entire image indicated by the drawing command group into regions based on the RGB color information, and performs different color conversions for each divided region.
[0031]
First, the printer correction processing unit 202 generates a raster image on an RGB 24-bit page memory by rasterizing a drawing command group based on the color information subjected to the image correction processing by the image correction processing unit 201. Then, the printer correction processing unit 202 generates CMYK data depending on the color reproducibility of the printer 106 for each pixel constituting the raster image, and transfers this to the printer 106.
[0032]
<Overall Operation of Image Correction Processing Unit 201>
FIG. 3 is a flowchart showing the overall operation of the image correction processing unit 201.
[0033]
Step S301:
The image correction processing unit 201 generates an area table by pre-scanning the entire image (hereinafter, also referred to as an “original image”) indicated by a drawing command group from the OS 102.
[0034]
Specifically, for example, the image correction processing unit 201 collects information on the entire original image by pre-scanning the original image (FIG. 5), divides the original image into regions based on this information, Generate Here, the area table corresponds to the original image (FIG. 5), for example, “two-dimensional coordinates” in which flags indicating “green area”, “sky area”, “skin area” and the like are described. This is a table (FIG. 8) in the form of "value-area flag".
[0035]
At the time of the pre-scan processing, the image correction processing unit 201 does not process all pixels constituting the original image, but samples, for example, 640 pixels × 480 pixels over the entire original image, and processes the sampled image. (FIG. 6) is processed.
By reducing the number of pixels to be processed in this way, it is possible to reduce the amount of memory used during the area division processing and to speed up the processing.
[0036]
Step S302:
The image correction processing unit 201 refers to the area table (FIG. 8) generated in step S301 and, based on this, converts each area (each raster) into which the original image has been divided into image conversions (corresponding to the divided areas). Different image processing is performed for each divided area.
[0037]
<Area Table Setting Process in Image Correction Processing Unit 201 (Step S301)> FIG. 4 shows a configuration for performing the process of step S301 shown in FIG. 3 in the image correction processing unit 201.
As shown in FIG. 4, the image correction processing unit 201 includes an image input unit 401, an image sampling unit 402, an image signal conversion unit 403, an image resampling unit 404, edge extraction units 405 and 406, an area flag setting unit 407, A boundary area selection unit 408 and a boundary area flag setting unit 409 are provided.
[0038]
Particularly, in the present embodiment, in the configuration shown in FIG. 4 described above, the edge extraction units 405 and 406 extract the edge of input image data (original image data) sampled by the image sampling unit 402 and the image resampling unit 404. The region flag setting unit 407 divides the region based on the spatial frequency information obtained by edge extraction and the color information of the input image. At this time, for an area that cannot be accurately divided from the sampled image data, as will be described later in detail, the area is divided at the actual resolution at the time of image correction, and the image correction processing corresponding to the divided area is performed.
[0039]
More specifically, first, the image input unit 401 receives an image constituted by a drawing command group from the OS 102, for example, image data 500 in RGB format as shown in FIG. Enter for. Here, the image size of the input image in FIG. 5 is, for example, data of 2560 pixels × 1920 pixels.
[0040]
The image sampling unit 402 samples the input image data 500 from the image input unit 401, and converts the image data 501 thinned out to a pixel number of about 640 pixels × 480 pixels, for example, a thumbnail, as shown in FIG. Generate.
In FIG. 6, the original image (FIG. 5) is schematically drawn in which sampling intervals are described. However, in actuality, the pixel values of one cell in FIG. 6 are the same as those of the original image (FIG. 5). It is the average value of the pixel values in the corresponding area.
The image signal conversion unit 403 converts the image data 501 obtained by the image sampling unit 402 into YCrCb format image data.
[0041]
The image resampling unit 404 samples the image data converted by the image signal converting unit 403 in the same manner as the image sampling unit 402, and generates, for example, image data 502 as shown in FIG. Here, FIG. 7 shows 320 pixels × 240 pixels.
The edge extraction unit 405 generates a low-frequency edge image by extracting a low-frequency region from the image data 502 obtained by the image resampling unit 404.
[0042]
The edge extracting unit 406 generates a medium frequency edge image by extracting a medium frequency region from the image data converted by the image signal converting unit 403.
[0043]
For the area flag setting unit 407, the low-frequency edge image obtained by the edge extraction unit 405, the medium-frequency edge image obtained by the edge extraction unit 406, and the image converted by the image signal conversion unit 403 (no conversion) 640 pixels × 480 pixels of a YCrCb format thinned image).
[0044]
The area flag setting unit 407, based on the input image shown in FIG. 6, may be configured based on the input image shown in FIG. 6, such as a specific color area such as a green area, a skin area, and a sky area (see FIG. Then, flags corresponding to these specific color areas are indicated by different types of oblique lines), and a borderless flag table including the border flags is generated.
[0045]
By the way, even if the area is divided on the thinned image obtained by roughly thinning the image, the boundary area on the thinned image shown in FIG. 8 is the boundary area of the original image (real resolution image) 500 shown in FIG. Does not correspond exactly. That is, since the resolution of the area table of FIG. 8 is set to be lower than that of the original image of FIG. 5, a fine structure in the original image (FIG. 5) cannot be expressed by the resolution of the area table of FIG. When an object is erroneously determined or an area to be divided into two at the resolution of the original image (FIG. 5) is recognized as one of the areas at the resolution of the area table in FIG. There is.
[0046]
Therefore, in the present embodiment, in this processing (the processing of step S301 shown in FIG. 3), coarse area division is performed as shown in FIG. Re-area division is performed at a resolution of (actual resolution image) 500.
[0047]
Specifically, the boundary area selection unit 408 selects, for example, an area indicated by “） (white circle)” as illustrated in FIG. 9 based on the area table without the boundary set by the area flag setting unit 407. .
In the selection of the boundary region here, for example, as shown in FIG. 10, when eight or more neighboring regions include a region different from the attention point by three or more of the eight neighboring pixels, the region is regarded as a boundary region. To do.
[0048]
For a region selected by the boundary region selection unit 408, for example, a region that is not shaded as shown in FIG. 11, a boundary region flag is set by the boundary region flag setting unit 409 as shown in FIG. .
[0049]
Therefore, in the setting of the area table in step 301 shown in FIG. 3, for example, a specific color area such as a green area, a skin area, and a sky area, a boundary area, and a non-specific color area excluding these are divided. Thus, an area table is generated for an image having a reduced image size of about a thumbnail.
[0050]
The area flag setting unit 407 sets an area table without boundaries, for example, with a configuration as shown in FIG.
[0051]
Specifically, first, as described above, the area flag setting unit 407 includes the low-frequency edge image obtained by the edge extraction unit 405, the medium-frequency edge image obtained by the edge extraction unit 406, and the image signal conversion. An image after conversion by the unit 403 (a non-converted YCrCb format thinned image of 640 pixels × 480 pixels) is input, and the area flag setting unit 407 sets a flag for area identification based on these input images. To generate an area table and output it.
[0052]
Therefore, as shown in FIG. 13, the area flag setting unit 407 includes a frequency determination unit 610 including a low frequency range determination unit 611 and a medium frequency range determination unit 612, a hue determination unit 621, a brightness determination unit 622, and a color determination unit. A chromaticity determination unit 620 including a degree determination unit 623, an area determination unit 630, and an area flag setting unit 640 are provided.
[0053]
Here, for simplicity of explanation, for example, the thinned image is an image of YCrCb data having a size of 640 pixels × 480 pixels, and an edge image in a low frequency region (low frequency edge image) is converted to a Y image having a size of 320 pixels × 240 pixels. Is a grayscale image processed by a Laplacian filter, and an edge image in a medium frequency region (middle frequency edge image) is a grayscale image obtained by processing Y data of 640 pixels × 480 pixels by a Laplacian filter.
[0054]
In the area flag setting unit 407, first, the low-frequency edge image is input to the low-frequency area determination unit 611, the medium-frequency edge image is input to the medium-low frequency area determination unit 612, and the thinned image is It is input to the chromaticity determination unit 620.
[0055]
The low frequency range determining unit 611, the middle and low frequency range determining unit 612, the hue determining unit 621, the lightness determining unit 622, and the saturation determining unit 623 included in the chromaticity determining unit 620 are each, for example, as illustrated in FIG. The region of the input image is divided according to a simple division method (quantitative division condition).
The region determining unit 630 specifies a corresponding region based on a region division result by the low frequency range determining unit 611, the middle and low frequency range determining unit 612, the hue determining unit 621, the lightness determining unit 622, and the saturation determining unit 623. I do.
[0056]
Note that areas that do not correspond to any area are classified as “other areas”.
[0057]
The area flag setting unit 640 sets a flag of an area corresponding to a corresponding portion on the area table based on the area specified by the area determining unit 630, and outputs this as an area table without boundaries.
[0058]
The quantitative division conditions are not limited to those shown in FIG. 14, but may be, for example, the conditions shown in FIG.
In the division conditions shown in FIG. 15, “Laplacian 320 × 240” indicates a result of processing a Y image resampled to 320 pixels × 240 pixels by a Laplacian filter, and “Laplacian 640 × 480” indicates 640 pixels × 480 pixels. Shows the processed Y image.
Further, in the division condition shown in FIG. 15, the processing is performed according to the threshold values for all the feature amounts. For example, the processing is performed according to a stochastic determination result as shown in FIG. You may comprise. In this case, the determination result may be reflected in the weighting vector using the probability of each area.
[0059]
As described above, in the present embodiment, in addition to the conventional configuration of region division using chromaticity, a condition in the frequency domain is used as the condition of division, so that the accuracy of region division is improved. be able to. In addition, by performing processing using a thinned image collected at the time of pre-scanning an input image in accordance with a condition in the frequency domain, it is possible to perform filtering processing in the real space domain at high speed.
[0060]
<Image conversion processing for each area in the image correction processing unit 201 (step S302)>
FIG. 17 shows a configuration for performing the process of step S302 shown in FIG. 3 in the image correction processing unit 201.
As shown in FIG. 17, the image correction processing unit 201 includes a selector 410, a local area flag setting unit 411, and a local area image processing determination unit 412 together with the image input unit 401 and the image signal conversion unit 403 shown in FIG. , An area-based image processing determination unit 413, a boundary area flag selection unit 414, a selector 415, an image conversion unit 416, an image signal conversion unit 417, and an image output unit 418.
[0061]
In the configuration shown in FIG. 17, the area table data obtained by the configuration shown in FIG. 4 (the configuration for performing the area table setting process in step S301 shown in FIG. 3) at the time of pre-scanning the original image , A different image conversion is performed for each corresponding image area.
[0062]
The present invention, in order to increase the accuracy of the region division, in the division by color and in the frequency domain,
1. Coarse region segmentation using low resolution thumbnail images
2. Region re-segmentation at the same (or close) resolution as the original image (only the part that cannot be determined by 1)
As described above, by performing the two-stage region division, it is possible to perform the region division that is light in processing and requires a small memory capacity.
[0063]
The area table data used here is 640 pixels × 480 pixels obtained by sampling the input image while the size of the input image to be processed (FIG. 5) is 2560 pixels × 1920 pixels. The size is relatively small (coarse in resolution).
[0064]
Therefore, although it can be expressed with an image size of 2560 pixels × 1920 pixels, if there is a fine structure (edge or the like) whose shape is destroyed when sampling is performed at 640 pixels × 480 pixels, the original image will be fine. Even if the color changes between the right and left sides of a simple structure (edge, etc.), as a result of the area division with coarse resolution, the left and right sides of the above structure (edge, etc.) are converted as the same color area. In some cases, it becomes impossible to express a detailed structure.
[0065]
Therefore, in the present invention, the region is roughly divided at a lower resolution than the original image, and the boundary of the divided region in the low-resolution region division is re-divided at a higher resolution immediately before the image conversion. In the conversion of the high-resolution original image, unnatural conversion is not performed on the high-resolution original image.
[0066]
Specifically, first, the image input unit 401 inputs image data (image data in RGB format) 500 as shown in FIG.
The image data input by the image input unit 401 is processed separately into “if not a boundary area” and “if it is a boundary area” as described below.
[0067]
If it is not a boundary area:
The image signal conversion unit 403 converts the RGB image data input by the image input unit 401 into YCrCb format image data.
[0068]
The image processing determination unit 413 refers to the area table obtained by the area table setting processing in step S301 for the image data converted by the image signal conversion unit 403, and determines whether the current conversion target pixel is a boundary area. As a result, if it is not a boundary area, based on the flag (area identification number) set in the area table, for example, green area color processing, skin area color processing, sky area color processing, A weight vector Wunit for the specific color area color processing is determined.
[0069]
Specifically, for example, as for the area G0 on the image shown in FIG. 5, a flag (area identification number) of a green area is set in the area table as shown in the “G0” part of FIG.
[0070]
Therefore, the image processing determining unit 413 determines, based on the above setting,
Wunit = {1 (green area), 0 (skin area), 0 (empty area), 0 (non-specific area)}
A color processing weight vector Wunit is set.
[0071]
The selector 415 selects the weighting vector Wunit set by the image processing determination unit 413 based on the signal indicating the non-boundary region supplied from the boundary region flag selection unit 414 when it is not the boundary region, and sends it to the image conversion unit 416. Supply.
The image conversion unit 416 performs image conversion on the image data converted by the image signal conversion unit 403 by weighting with the weight vector Wunit from the selector 415.
[0072]
The image data (YCrCb format image data) converted by the image conversion unit 416 is converted into RGB format image data by the image signal conversion unit 417, and output via the image output unit 418.
[0073]
If it is a boundary area:
The image signal conversion unit 403 converts the RGB image data input by the image input unit 401 into YCrCb format image data.
[0074]
The image processing determination unit 412 refers to the area table obtained in the area table setting process in step S301 for the image data converted by the image signal conversion unit 403, and determines whether the current conversion target pixel is a boundary area. Is checked, and as a result, when it is recognized that the area is a boundary area (see FIG. 12), image processing is sequentially determined for each pixel.
[0075]
Specifically, for example, in the present embodiment, in the boundary region of the image defined by the region table (FIG. 8) having a resolution lower than that of the original image (FIG. 5), the image processing changes before and after the boundary. In order to prevent the discontinuity of the gradation, the local area image processing determination unit 412 applies a load to each of the image processing in the nearby area in the thinned image and the image processing in the area set by the local area flag setting unit 411. An average is obtained, and a weighting vector Wlocal for color processing is generated and output based on the average.
[0076]
The selector 415 selects the weighting vector Wlocal set by the image processing determination unit 412 based on the signal indicating the boundary region supplied from the boundary region flag selection unit 414 when the image is a boundary region, and sends the weighted vector Wlocal to the image conversion unit 416. Supply.
The image conversion unit 416 performs image conversion on the image data converted by the image signal conversion unit 403 by weighting with the weight vector Wlocal from the selector 415.
[0077]
The image data (YCrCb format image data) converted by the image conversion unit 416 is converted into RGB format image data by the image signal conversion unit 417, and output via the image output unit 418.
[0078]
FIG. 18 specifically shows the configuration of the local area flag setting unit 411 shown in FIG.
As shown in FIG. 18, the local area flag setting section 411 includes a hue determining section 701, a lightness determining section 702, a saturation determining section 703, a local area determining section 704, and a local area flag setting section 705.
[0079]
When the region to be processed is a boundary region, image data in YCrCb format is input to the local region flag setting unit 411 via the selector 410. The local area flag setting unit 411 uses the configuration illustrated in FIG. 18 to determine to which area the processing target area locally belongs in the input image data.
[0080]
That is, in the local area flag setting unit 411, the hue determination unit 701, the lightness determination unit 702, and the saturation determination unit 703 determine the input image data similarly to the area flag setting unit 407 illustrated in FIGS. The corresponding area is determined.
[0081]
The local area determination unit 704 determines a corresponding area based on the determination results by the hue determination unit 701, lightness determination unit 702, and saturation determination unit 703. The determination here is made based on items other than the frequency domain in FIG.
The local area flag setting unit 705 outputs an identification flag (area identification flag) of the corresponding area based on the area determined by the local area determination unit 704.
[0082]
Here, as an example, the determination in the spatial frequency domain using the filter processing is performed using only the chromaticity because the calculation amount increases when the determination is performed on the real resolution image. For example, when the configuration of the present embodiment is realized by an apparatus or system capable of coping with an increase in the amount of calculation, it is needless to say that the determination in the spatial frequency domain for the real resolution image may be used together.
Further, as an example, similar to the area flag setting unit 407 illustrated in FIGS. 4 and 13, the processing is performed by using a threshold, and a scalar flag (area identification flag) is output. A probabilistic determination as shown in FIG. 16 may be performed, a flag in the form of a vector may be output, and this may be reflected in the weight vector Wlocal.
[0083]
FIG. 19 specifically shows the configuration of the image conversion unit 416 shown in FIG.
The image conversion unit 416 performs image conversion of the input image data (original image data) by a process weighted by a weight vector Wlocal for color processing. As shown in FIG. 19, the LUT calculation units 801 to 803 and 809 are used. , Weighting processing units 804 to 807, and an adder 808.
[0084]
The weighting vector W of each color process selected by the selector 415 (see FIG. 17) and the image data converted into the YCrCb format by the image signal converting unit 403 (see FIG. 17) are input to the image converting unit 416. You.
[0085]
Each of the LUT calculation units 801 to 803 and 809 has a database (not shown) in which a color processing table for a corresponding color area is stored, and converts the input image data using the color processing table.
Each of the weighting processing units 804 to 807 weights and outputs the image data supplied from the corresponding calculation unit in the LUT calculation units 801 to 803 and 809 according to the weight vector W of the corresponding color processing.
The adder 808 adds the outputs of the weighting processing units 804 to 807, and outputs the addition result as converted image data in the YCrCb format.
[0086]
The database (DB) of each of the LUT calculation units 801 to 803 and 809 includes, for example, a color conversion table optimized for a green region, a color conversion table optimized for a sky blue region, and a color conversion table for a skin color region. An optimized color conversion table and a general-purpose color conversion table for other areas are stored. As a format of these color conversion tables (LUT), for example, RGB values after conversion are obtained for grid points of R, G, B = {0, 32, 64, 96, 128, 160, 192, 224, 255}. The format of the recorded “3D LUT” is used. Therefore, the LUT calculation units 801 to 803 and 809 process the input image data by an interpolation calculation using a known interpolation technique such as tetrahedral interpolation using a “3D LUT” color conversion table. , And converts the RGB values of the input image data.
[0087]
FIG. 20 specifically shows the configuration of the local area image processing determination unit 412 shown in FIG.
The local area image processing determination unit 412 generates and outputs a weight vector Wlocal for color processing based on data from the area table and an area identification flag corresponding to the local area set by the local area flag setting unit 411. 20. As shown in FIG. 20, a proximity area flag detection unit 901 and an area flag combination unit 902 are provided.
[0088]
The neighborhood area flag detection unit 901 checks the area identification flag of the neighborhood area when viewed at a low resolution of 640 pixels × 480 pixels lower than the actual image resolution, and counts each area.
The area flag synthesizing unit 902 includes an area identification flag obtained by the local area flag setting unit 411 and an area identification flag obtained by the adjacent area flag detection unit 901 in order to suppress the occurrence of discontinuous gradation before and after the boundary. By weighting and combining the area identification flag according to the count value (count) of each area identification flag, a weight vector Wlocal of each area is generated and output.
[0089]
FIG. 21 shows the configuration of the local area image processing determination unit 412 shown in FIG. 20 in more detail.
[0090]
As shown in FIG. 21, the local area image processing determination unit 412 includes a skin area coefficient unit 911, a sky area coefficient unit 912, a green area coefficient unit 913, other area coefficient units 914, an adder 915, a counter 916, and a divider. 917 and a vectorization unit 918.
[0091]
With the configuration shown in FIG. 21 described above, the local area image processing determination unit 412 determines the consistency between the area information determined based on the local image information and the surroundings for the area that is the boundary area on the area table. In this way, it is possible to prevent the discontinuity of the gradation while performing the precise area division.
[0092]
The local area image processing determination unit 412 outputs an area weighting vector Wlocal based on the local flag set by the local area flag setting unit 411 and the area near the target pixel to be processed in the area table.
[0093]
The weighting vector Wlocal is obtained by calculating the skin, sky, green, and other color regions in the region near the target pixel to be processed in the region table, the respective counting results (the counting results of the counting units 911 to 914), and the local region flag setting unit. The divider 917 normalizes the sum so that the sum with the local flag set in 411 becomes “1”, and outputs this as a unit vector (Wlocal) by the vectorization unit 918.
[0094]
In the present embodiment, different color processing is performed for each divided region. However, the present invention is not limited to this. For example, by performing different filtering for each divided region, It is also possible to produce an effect like a soft focus filter used for blurring the background.
[0095]
Further, an object of the present invention is to provide a storage medium storing program codes of software for realizing the functions of the present embodiment to a system or an apparatus, and a computer (or CPU or MPU) of the system or the apparatus to execute the storage medium. Needless to say, this can also be achieved by reading and executing the program code stored in the.
In this case, the program code itself read from the storage medium realizes the function of the present embodiment, and the storage medium storing the program code and the program code constitute the present invention.
As a storage medium for supplying the program code, a ROM, a flexible disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a CD-R, a magnetic tape, a nonvolatile memory card, or the like can be used.
The functions of the present embodiment are realized not only by executing the program code read by the computer, but also by the operating system running on the computer based on the instruction of the program code. It goes without saying that a part or all of the above is performed and the function of the present embodiment is realized by the processing.
Further, after the program code read from the storage medium is written into a memory provided in an extension function board inserted into the computer or a function extension unit connected to the computer, the function extension is performed based on the instruction of the program code. It goes without saying that the CPU or the like provided in the board or the function expansion unit performs part or all of the actual processing, and the processing realizes the functions of the present embodiment.
[0096]
【The invention's effect】
As described above, in the present invention, when a target image is divided into regions and color processing (correction processing) is performed on the divided regions, the target image is divided based on color information and spatial frequency information obtained from the target image. It was configured to divide the area.
As a result, even if the color processing is performed for each area after the area is divided, a pseudo contour or the like does not occur at the area switching portion, and a good image after the color processing can be provided.
[0097]
In addition, if a low-resolution image obtained by sampling the target image is configured to be a target of region division, simple region division requiring a small amount of memory and a small amount of calculation can be performed.
[Brief description of the drawings]
FIG. 1 is a block diagram illustrating a configuration of a host computer to which the present invention has been applied.
FIG. 2 is a block diagram illustrating a configuration of a printer driver.
FIG. 3 is a flowchart for explaining processing of image information collection and image conversion.
FIG. 4 is a block diagram illustrating a configuration of an image conversion unit.
FIG. 5 is a diagram for explaining an example of an image (original image) to be processed.
FIG. 6 is a diagram illustrating an example of a thinned-out image having a size of 640 pixels × 480 pixels from which image information is collected in pre-scanning of an original image.
FIG. 7 is a diagram for explaining an example of a low-resolution image obtained by re-sampling a thinned image having a size of 640 pixels × 480 pixels.
FIG. 8 is a diagram for explaining an example of a borderless flag table including border flags to which only flags corresponding to specific color areas are assigned.
FIG. 9 is a diagram for explaining an area selected as a boundary area in the no-boundary flag table.
FIG. 10 is a diagram for explaining eight neighborhoods of a target pixel.
FIG. 11 is a diagram for explaining a correspondence between a boundary region and an original image.
FIG. 12 is a diagram illustrating an area table in which a boundary area flag is set.
FIG. 13 is a block diagram illustrating a configuration of an area flag setting unit.
FIG. 14 is a diagram for describing a qualitative area division criterion.
FIG. 15 is a diagram for describing a criterion for quantitative area division.
FIG. 16 is a diagram illustrating a variation of a processing method based on a threshold value for a feature amount.
FIG. 17 is a block diagram illustrating a configuration of an image correction processing unit.
FIG. 18 is a block diagram illustrating a configuration of a local area flag setting unit.
FIG. 19 is a block diagram illustrating a configuration of an image conversion unit.
FIG. 20 is a block diagram illustrating a configuration of a local area image processing determination unit.
FIG. 21 is a block diagram illustrating a configuration of a local area image processing determination unit in detail.
FIG. 22 is a diagram illustrating color processing performed by a printer driver.
[Explanation of symbols]
100 Host computer
101 Application
102 OS
103 Printer Driver
104 monitor driver
105 monitor
106 Printer
107 HD
108 CPU
109 RAM
110 ROM
201 Image correction processing unit
401 Image input unit
402 Image sample part
403 Image signal converter
404 Image resampling unit
405 Edge extraction unit (low frequency)
406 Edge extractor (medium frequency)
407 area flag setting unit
408 Boundary area selector
409 Boundary area flag setting unit

Claims (18)

An image processing apparatus that divides a target image into regions and performs color processing on the divided regions,
An image processing apparatus comprising: a dividing unit that divides a target image into regions based on color information and spatial frequency information obtained from the target image. The image processing apparatus according to claim 1, wherein the spatial frequency information includes information obtained from a low-resolution image obtained by sampling a target image. The dividing means divides the target image into regions based on the spatial frequency information obtained from the low-resolution image obtained by sampling the target image. For the region that cannot be divided by the region division, the spatial frequency information obtained from the real-resolution target image is used. 2. The image processing apparatus according to claim 1, wherein the target image is divided into regions based on the following. 2. The image processing apparatus according to claim 1, further comprising processing means for performing color processing according to the target area for each of the divided areas obtained by the dividing means. An image processing apparatus that performs a correction process on a target image based on characteristics of the target image detected by pre-scanning the target image,
Dividing means for performing area division of the target image input at the time of the prescan, and for an area not determined by the area division, performing area division again at the time of the correction processing;
An image processing apparatus comprising: a processing unit that performs the correction process according to a target region for each of the divided regions obtained by the dividing unit. 6. The image processing apparatus according to claim 5, wherein the dividing unit divides the target image into regions based on color information and spatial frequency information obtained from the target image. 6. The image according to claim 5, wherein the division unit divides an image having a lower resolution than the input image obtained by sampling the input image in the area division of the target image input at the time of the pre-scan. Processing equipment. An image processing system in which a plurality of devices are communicably connected to each other,
An image processing system, wherein at least one of the plurality of devices has a function of the image processing apparatus according to claim 1. An image processing method for dividing a target image into regions so that similar colors are the same,
An input step of inputting a target image;
A color information recognition step of recognizing color information of the target image input in the input step;
A spatial frequency information recognition step of recognizing spatial frequency information of the target image input by the input step,
An image region recognition step of recognizing a target image region based on the color information recognized in the color information recognition step and the spatial frequency information recognized in the spatial frequency information recognition step. The image processing method according to claim 9, wherein the step of recognizing the spatial frequency information includes a step of recognizing spatial frequency information of a low-resolution image obtained by sampling the target image. An image processing method for pre-scanning the target image to detect characteristics of the target image, and performing a correction process on the target image based on the detection result, wherein the input image is divided into regions at the time of the pre-scan, An image processing method comprising: performing a region division again in the correction process for an area not determined by the region division at the time of the prescan, and performing the correction process according to the divided region. Method. The above processing steps are steps for performing region division so that similar colors of the target image are the same.
An input step of inputting a target image;
A color information recognition step of recognizing color information of the target image input in the input step;
A spatial frequency information recognition step of recognizing spatial frequency information of the target image input by the input step,
The image processing apparatus according to claim 11, further comprising: an image area recognition step of recognizing a target image area based on the color information recognized in the color information recognition step and the spatial frequency information recognized in the spatial frequency information recognition step. Image processing method. The image processing method according to claim 11, wherein the processing step includes a step of dividing an image having a lower resolution than the target image obtained by sampling the target image in the region division at the time of the pre-scan. A program for causing a computer to function as predetermined means,
The above-mentioned predetermined means is a device or system that divides the target image into regions and performs color processing on the divided regions,
A program comprising means for dividing a target image into regions based on color information and spatial frequency information obtained from the target image. A program for causing a computer to function as predetermined means,
The predetermined means is a means provided in an apparatus or a system for performing a correction process on the target image based on characteristics of the target image detected by pre-scanning the target image,
Dividing means for performing area division of the target image input at the time of the prescan, and for an area not determined by the area division, performing area division again at the time of the correction processing;
A program, comprising: processing means for performing the correction processing according to a target area for each divided area obtained by the dividing means. A program for causing a computer to execute predetermined steps,
The predetermined step is a processing step for performing area division such that similar colors of the target image are the same,
An input step of inputting a target image;
A color information recognition step of recognizing color information of the target image input in the input step;
A spatial frequency information recognition step of recognizing spatial frequency information of the target image input by the input step,
A program for recognizing a target image region based on the color information recognized in the color information recognition step and the spatial frequency information recognized in the spatial frequency information recognition step. A program for causing a computer to execute predetermined steps,
The predetermined step is a processing step for pre-scanning the target image to detect a feature of the target image, and performing a correction process on the target image based on the detection result.
The input image is divided into regions at the time of the pre-scan, and the region not determined by the region division at the time of the pre-scan is subjected to the region division again at the time of the correction processing, and the correction according to the divided region is performed. A program comprising processing steps for performing processing. A computer-readable storage medium on which the program according to claim 14 is recorded.

Images