JP2008227897A - Unit and method for processing image - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To output an image easily by applying correction to a desired image in the same way. <P>SOLUTION: Deviation of color used in a reference image is detected for each inspection region, in a plurality of inspection regions set in a color space. Based on the deviation of use detected for each inspection region, the amount of correction applied to the reference image is estimated, and image data are corrected based on the estimated amount of correction, thus estimating the amount of correction applied to the reference image, even if there are no uncorrected image data, and hence, correcting user's image data to have similar impressions as those of the reference image for output. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to a technique for correcting image data.

  Nowadays, images are often handled in the form of digital data (that is, image data). Accordingly, when image data is output as an image, it is not output as it is, but a more favorable impression. In general, the image is output after being subjected to some correction processing so as to be an image. In recent years, the content of correction processing applied to image data when outputting an image is important in determining the impression of the image along with the composition of the image and the drawn (or photographed) object. It is an element.

  In the present specification, the correction processing applied to the image data in order to obtain a preferable impression image is referred to as “picture making”. When outputting image data, for example, some conversion is usually performed on the image data in order to absorb differences in the characteristics of the device. It is a broad concept that includes all image transformations that are performed to produce a more favorable image beyond the transformations that are performed to correctly reproduce the data.

  Despite the fact that picture making is an important factor that determines the impression of an image, advanced expertise and experience are required to make an appropriate picture. It is difficult to make a simple picture unless you are an expert such as a professional photographer or creator. Therefore, memorize the contents of the picture making performed by various experts on various images, select an image close to the image to be output, and make the picture making with the contents stored for the selected image. A technique for performing the above has been proposed (Patent Document 1). In addition, a technique for accurately extracting the contents of picture making performed by an expert has been proposed (Patent Document 2).

JP 2006-005766 A Japanese Patent No. 2001-045316

  However, with the proposed technology, it was still difficult for general users to freely create their favorite pictures. This is because in the proposed technology, the contents of the picture making performed by the expert are stored in advance, or in order to extract the contents of the picture making of the expert, the image data before the picture making and the after the picture making It is assumed that the image data can be obtained. In reality, however, the content of painting is often a know-how element, and it is rare that a large number of painting content is stored in a selectable manner, and image data before and after painting can be obtained. Is rare. For this reason, there is a problem that it is still difficult for a general user to make a favorite picture and output an image.

  The present invention has been made in order to solve the above-described problems of the prior art, and an object of the present invention is to provide a technique that allows a general user to easily output an image of a favorite picture creation.

In order to solve at least a part of the problems described above, the image processing apparatus of the present invention employs the following configuration. That is,
An image processing apparatus that receives image data and applies correction,
Inspection area setting means for setting a plurality of inspection areas in a predetermined color space;
Analyzing reference image data of a reference image selected in advance, a reference image analysis means for detecting a bias in use frequency of colors in the inspection region for each inspection region;
A correction amount determining means for determining a correction amount to be added to the image data based on a bias in use frequency detected for each inspection region;
And a correction execution means for executing correction of the received image data based on the determined correction amount.

The image processing method of the present invention corresponding to the above image processing apparatus is
An image processing method for receiving and correcting image data,
A step of setting a plurality of inspection areas in a predetermined color space;
Analyzing the pre-selected reference image image data to detect a bias in use frequency of colors in the inspection region for each inspection region;
Determining a correction amount to be added to the image data based on a bias in use frequency detected for each inspection area;
And a step of correcting the received image data based on the determined correction amount.

  In the image processing apparatus and the image processing method according to the present invention, a deviation in the usage frequency of the colors used in the reference image is detected for each inspection region within a plurality of inspection regions set in the color space. Then, the correction amount added to the reference image is estimated based on the bias in the usage frequency detected for each inspection region, and the image data is corrected based on the estimated correction amount correction amount.

  If the correction amount of the reference image is estimated in this way, the correction amount added to the reference image can be estimated even if there is no image data before correction. Normally, even if a reference image corrected to a favorable impression is found, if there is no image data before correction, it is not known how it is corrected, so a similar impression can be obtained with the image data on hand. However, in the image processing apparatus and the image processing apparatus of the present invention, it is possible to correct the image data on hand to an image having the same impression as the reference image.

  Further, in the image processing apparatus of the present invention, with respect to a plurality of inspection areas set with at least different hues in the color space, the use frequency bias of colors used in the reference image is determined for each inspection area. After the detection, the correction amount for the hue may be determined based on the detected bias in the usage frequency, and the image data may be corrected. Here, the color space may be any color space as long as the three elements of lightness, hue, and saturation can be expressed directly or indirectly, for example, RGB color space, Lab color, and the like. Space, CMY color space, HSV color space, HSL color space, YCC color space Luv color space, XYZ color space, and the like can be used.

  Correcting the hue greatly affects the impression of the image. Therefore, if the correction amount is estimated based on a plurality of inspection areas set with different hues, the image data is corrected so that a favorable impression is obtained. It becomes possible.

  In the image processing apparatus of the present invention, the correction amount may be determined as follows. First, as the inspection area, inspection areas for the red (R), green (G), and blue (B) hues corresponding to the three primary colors of light are set. Further, the image data of the reference image is received in the form of RGB image data represented by RGB components, and the frequency distribution in the inspection region of each hue of RGB is detected. Then, the correction amount for the R hue is determined based on the deviation between the frequency distribution detected in the R inspection region and the R component of the RGB image data. Similarly, based on the deviation between the frequency distribution detected in the G inspection area and the G component of the RGB image data, a correction amount for the G hue is determined, and the frequency distribution in the B inspection area and the RGB image are determined. The correction amount for the hue of B may be determined from the deviation from the B component of the data.

  Since image data is usually often expressed by RGB components, it is preferable to obtain the correction amount of image data for each RGB component because the image data can be easily corrected.

Furthermore, the present invention can be realized using a computer by causing a computer to read a program for realizing the above-described image processing method and executing a predetermined function. Therefore, the present invention also includes the following aspects as a program. That is, the program of the present invention corresponding to the image processing method described above is
A program for realizing an image processing method for receiving and correcting image data using a computer,
A function for setting a plurality of inspection areas in a predetermined color space;
A function of detecting a bias in use frequency of colors in the inspection area by analyzing image data of a reference image selected in advance, for each inspection area;
A function of determining a correction amount to be added to the image data based on a bias in use frequency detected for each inspection region;
The gist of the present invention is that a function of executing correction of the received image data is realized by a computer based on the determined correction amount.

  If such a program is read into a computer and each of the above functions is realized, even a general user can easily output a favorite picture-making image.

Hereinafter, in order to clarify the contents of the present invention described above, examples will be described in the following order.
A. Device configuration:
B. Image output processing:
C. Variation:

A. Device configuration :
FIG. 1 is a perspective view showing an external shape of a printing apparatus 10 equipped with the image processing apparatus of this embodiment. As illustrated, the printing apparatus 10 according to the present exemplary embodiment includes a scanner unit 100, a printer unit 200, a control unit 300 that controls operations of the scanner unit 100 and the printer unit 200, and the like. The scanner unit 100 has a scanner function of reading a printed image and generating image data, and the printer unit 200 has a printer function of receiving image data and printing an image on a print medium. . The control unit 300 is a CPU, a ROM, a RAM, a D / A converter that converts digital data into an analog signal, and a peripheral device interface PIF for exchanging data with peripheral devices. And has a function of controlling operations of the scanner unit 100 and the printer unit 200. Further, the control unit 300 can perform predetermined image processing on image data of an image read by the scanner unit 100 or image data supplied from the outside, and output an image from the printer unit 200. .

B. Image output processing:
FIG. 2 is a flowchart showing the flow of image output processing performed when the printing apparatus 10 outputs an image. Such a process is a process executed by the control unit 300 in the printing apparatus 10.

  As shown in the figure, in the image output process of this embodiment, first, image data of a reference image is acquired (step S100). Here, the reference image is an image selected in advance as a preferable image by a person who intends to output the image. The image data of the reference image may be acquired by reading the reference image with the scanner unit 100, or may be acquired by reading image data that has been acquired and stored in advance.

  Next, a histogram in the sampling color gamut is acquired by analyzing the image data of the acquired reference image (step S102). Here, the sampling color gamut is a color gamut preset in a color space (here, referred to as a Lab color space). In consideration of the fact that color image data is usually expressed by RGB color components, in this embodiment, a color gamut representing a red (R) color, a color gamut representing a green (G) color, and a blue (B) color. Sampling color gamuts are respectively set in the color gamuts representing Of course, the sampling color gamut is not limited to such a color gamut. For example, a cyan (C) color, a magenta (M) color, a yellow (Y) color gamut, a purple or orange color gamut, or the like. , It can be set to a color portion that affects the impression of the image. Note that it is desirable to set these sampling color gamuts avoiding regions where the lightness and saturation are too high and low.

  Also, the color space for setting the sampling color gamut is not limited to the Lab color space. For example, the RGB color space, CMY color space, HSV color space, HSL color space, YCC color space, Luv color space, XYZ It is also possible to set a sampling color gamut in various color spaces such as a color space.

  FIG. 3 conceptually shows how three sampling color gamuts for each color of RGB are set in the Lab color space. In this embodiment, histograms in these sampling color gamuts are acquired as follows. First, the image data of the reference image is obtained, and the color histogram used for the reference image is generated by plotting the color of each pixel in the Lab color space. Next, by extracting a portion of the sampling color gamut from the generated histogram, a histogram in each sampling color gamut is acquired. Of course, the image data of the reference image may be analyzed to generate a histogram only for the sampling color gamut from the beginning.

  FIG. 4 is an explanatory diagram conceptually showing an example of a histogram acquired in the sampling color gamut for R. In the example shown in the drawing, the red used for the reference image is often a slightly yellowish red. By acquiring histograms for the sampling color gamut for G and the sampling color gamut for B as well, what kind of green color is used in the reference image, and what kind of blue color is frequently used for blue I can know.

  In the image output process shown in FIG. 2, when a histogram is acquired for each sampling color gamut in this way, the histogram bias is acquired for each sampling color gamut (step S104). In this embodiment, the histogram bias is calculated by the following method. First, the coordinate value of the most frequently used color is detected within the sampling color gamut. In the example shown in FIG. 4, coordinate values corresponding to a slightly yellowish red are detected. Then, a deviation from a reference red coordinate value preset in the sampling color gamut is calculated. As the reference red color, the color at the center in the sampling color gamut can be used. Alternatively, the R color defined by the color image data standard can be used as the reference red color.

  Based on the bias of the histogram thus obtained for each sampling color gamut, what kind of correction is performed on the reference image, in other words, the amount of correction of the reference image is estimated (step S106). Here, the principle of estimating the correction amount added to the reference image from the bias of the histogram obtained for each sampling color gamut in the image output processing of the present embodiment will be described.

  In order to know the correction amount applied to the reference image, it is usually considered that the image data before correction and the image data after correction must be known. In view of this, even if a rough correction amount is estimated, it is considered that it is necessary to estimate the image data before correction in some form. However, if the viewpoint is changed as follows, it is possible to estimate a rough correction amount added to the reference image without estimating the image data before correction. For example, in the example shown in FIG. 4, the red color frequently used in the reference image is a slightly yellowish red color. Then, it can be considered that the most frequently used red color has a great influence on the impression of the reference image. Similarly, for green and blue, what kind of green is the most used green, or what is the most used blue is a big influence on the impression of the reference image. It is considered that these effects combine to determine the impression of the reference image. Therefore, for example, if the color is red, the image data is corrected so that the color is slightly yellowish red. The specific correction amount may be determined so that the standard red color set in advance in the sampling color gamut for R becomes a slightly yellowish red color obtained from the reference image.

  FIG. 5 conceptually shows how the correction amount for red is determined in this way. In the illustrated example, R defined in the sRGB standard, which is a standard for RGB image data, is used as the reference red color. Since the RGB image data of the sRGB standard is expressed by a combination of R, G, and B specified in the standard, the red color used in the image of the sRGB image data is commonly R based on the standard. It is thought that there are many cases where the distribution is centered around. Therefore, it is possible to approximate the impression of the reference image as a whole by correcting R of the sRGB standard so as to match the slightly yellowish red color obtained from the reference image. For other colors (green and blue), the correction amount determined in the same way and correcting each color will correct the image so that it approaches the impression of the reference image as a whole, although it is not accurate. It becomes possible to do.

  Also, the image data is usually output in a state in which the color deviation is corrected and the image is automatically corrected so as to obtain a more natural impression image, for example, when white balance is taken. is there. Similarly, when only a small part of the representable gradation range is used, it is usually output in a state in which correction for adjusting the contrast is automatically performed. As a result, unless the image data is corrected specifically, the histogram of the RGB image data naturally exhibits a certain tendency. Therefore, instead of simply using standard R, G, and B as standard colors for sRGB, each sampling color is considered by considering various corrections that are performed semi-automatically for these standard colors. It is also possible to set a reference color for the area. If the standard color set in this way is corrected so as to match the color characterizing the reference image obtained in each sampling color gamut, an image with an impression closer to the reference image can be obtained. It becomes possible.

  Furthermore, from the opposite viewpoint, this can be considered as follows. That is, unless a particularly intended correction is applied, the histogram of each sampling color gamut should not deviate significantly from a preset reference color. However, if a large deviation occurs from the reference color, it is considered that a correction close to the deviation amount is applied. In this way, rough correction contents can be estimated from only the image data of the corrected image (that is, the reference image).

  In step S106 of the image output process shown in FIG. 2, based on the principle as described above, it is added to the reference image as illustrated in FIG. 5 from the bias of the histogram detected in each sampling color gamut. Estimate the correction amount.

  In the above description, sRGB standard RGB or colors that take into account various corrections that are automatically performed on these standard RGB colors are used as the reference color, thereby being added to the reference image. The amount of correction was estimated. However, it is not necessarily limited to the RGB of the sRGB standard, and other standard RGB or RGB uniquely set by the user (further, these RGB colors that take various corrections into account) can be used as reference colors. . In addition, it is also possible for the user to set his / her favorite color in advance and use that color as the reference color.

  After the correction amount added to the reference image is estimated in this way, thereafter, the same processing as that for outputting a normal image may be performed. Briefly described below, image data to be output is acquired (step S108). Next, the image data is corrected using the estimated correction amount (step S110). For example, in the example shown in FIG. 5, for the red hue, the hue is slightly corrected in the direction of the yellow side. At this time, the correction amount may be gradually reduced so that color discontinuity does not occur at the boundary between the corrected region and the uncorrected region.

  Alternatively, the correction amount estimated from the reference image may not be used as it is, but the correction amount may be reduced at a certain ratio to correct the image data. Furthermore, the image data corrected using the estimated correction amount and the original image data that has not been corrected may be combined at a fixed ratio to obtain final corrected image data. In this way, if the estimated correction amount is reduced by a certain ratio, or if the corrected image data and the uncorrected image data are combined into final image data, the correction amount is consequently suppressed. Therefore, even if the estimated correction amount is inappropriate, it is possible to correct the image data without causing major breakdown.

  When the image data is corrected in this way, an image is output according to the corrected image data (step S112). The form of outputting an image may be, for example, a form in which ink dots are formed on a print medium and the image is printed, or a form in which toner is attached to the print medium and the image is printed. Alternatively, it is also possible to display an image with colored light on a monitor or screen instead of a print medium.

  As described above, in the image output process according to the present embodiment, the correction amount added to the reference image can be estimated without obtaining the image data before correction, and the corrected image can be output. . For this reason, it is possible to correct and output the image data on hand so that it becomes an image having an impression close to that image only by obtaining an image having a preferable impression, for example.

C. Modified example:
In the above-described embodiments, the image data is corrected in the hue direction. However, it is also possible to correct the image data in the saturation direction or the brightness direction. For example, the correction amount in the saturation direction may be estimated by setting a plurality of sampling color gamuts in the saturation direction instead of the hue direction, and obtaining a histogram in each sampling color gamut. Alternatively, as illustrated in FIG. 6, a plurality of sampling gamuts may be set by dividing the sampling gamut in the saturation direction for each hue. In this way, it is possible to estimate the correction amount in the saturation direction applied to the image from the reference image. Of course, it is also possible to estimate the correction amount in both the saturation direction and the hue direction.

  Although the printing apparatus of the present embodiment has been described above, the present invention is not limited to all the embodiments and modifications described above, and can be implemented in various modes without departing from the spirit of the present invention.

1 is a perspective view showing an external shape of a printing apparatus equipped with an image processing apparatus according to an embodiment. 6 is a flowchart illustrating a flow of image output processing performed when an image is output by the printing apparatus. It is explanatory drawing which showed notionally the mode that three sampling color gamuts for each color of RGB were set in the Lab color space. It is explanatory drawing which showed notionally an example of the histogram acquired by the sampling color gamut for R. It is explanatory drawing which showed notionally the mode that the correction amount about red was determined. It is explanatory drawing which illustrated a mode that the some sampling color gamut was set to the saturation direction.

Explanation of symbols

10 ... Printing device, 100 ... Scanner section,
200: Printer unit, 300: Control unit

Claims (5)

An image processing apparatus that receives image data and applies correction,
Inspection area setting means for setting a plurality of inspection areas in a predetermined color space;
Analyzing reference image data of a reference image selected in advance, a reference image analysis means for detecting a bias in use frequency of colors in the inspection region for each inspection region;
A correction amount determining means for determining a correction amount to be added to the image data based on a bias in use frequency detected for each inspection region;
An image processing apparatus comprising: a correction execution unit that executes correction of the received image data based on the determined correction amount. The image processing apparatus according to claim 1,
The inspection area setting means is a means for setting the plurality of inspection areas with at least different hues in the color space,
The correction amount determination means is an image processing apparatus which is a means for determining a correction amount for a hue based on the bias in use frequency detected for each inspection region. The image processing apparatus according to claim 2,
The inspection area setting means is a means for setting the inspection area for each hue of red (R), green (G), and blue (B) corresponding to the three primary colors of light,
The reference image analysis means is means for receiving and analyzing the image data of the reference image in the form of RGB image data represented by RGB components,
The correction amount determining unit is configured to correct a correction amount for each hue of RGB based on a deviation between the use frequency detected in the inspection region for each hue of R, G, and B and each RGB component of the RGB image data. An image processing apparatus as means for determining An image processing method for receiving and correcting image data,
A step of setting a plurality of inspection areas in a predetermined color space;
Analyzing the image data of the reference image selected in advance, detecting a bias in the use frequency of the colors in the inspection region for each inspection region;
Determining a correction amount to be added to the image data based on a bias in use frequency detected for each inspection area;
An image processing method comprising: correcting the received image data based on the determined correction amount. A program for realizing an image processing method for receiving and correcting image data using a computer,
A function for setting a plurality of inspection areas in a predetermined color space;
A function of detecting a bias in the use frequency of the colors in the inspection region for each inspection region by analyzing image data of a reference image selected in advance;
A function of determining a correction amount to be added to the image data based on a bias in use frequency detected for each inspection area;
The computer implement | achieves the function which performs correction | amendment of the received image data based on the determined correction amount.

Images