JP2007325134A - Image correction apparatus, image correction method, image correction program, image reading apparatus, image reading method, and printer - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technology of applying proper correction processing to an image the color balance of a halftone of which is unbalanced. <P>SOLUTION: A correction apparatus includes: a data generating section that generates data for each of histograms by each color for expressing the distribution of the number of pixels with respect to a density value of each color of pixels configuring an image; an attribute information acquisition section that respectively demarcates a region expressed by each histogram by each color into three or more small regions by the magnitude of the respective density values on the basis of data of each histogram by each color, selects at least one small region among the three or more small regions, and acquires the attribute information associated with the small region by each color; a target value acquisition section that selects a small region whose density value is greater than that of a small region and smaller than that of the other small region among the three or more small regions as a selected small region by each color and acquires a target value shared in each color as to the attribute information associated with the selected small region on the basis of the acquired attribute information; and a correction processing section for applying correction processing to the image so that the attribute information of each color associated with the selected small region respectively reach the target value shared in each color. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an image correction apparatus, an image correction method, an image correction program, an image reading apparatus, an image reading method, and a printing apparatus.

  Various image reading apparatuses such as an image scanner are generally connected to a so-called computer apparatus such as a personal computer by wire or wireless, and transmit an image read from a document as image data to the computer apparatus. The computer device receives the image data transmitted from the image reading device and stores it in a data storage device such as various memories or a hard disk device. At this time, the computer apparatus performs various adjustment processes on the image data sent from the image reading apparatus.

Examples of the adjustment processing performed here include histogram adjustment for adjusting the brightness of an image, density correction for partially changing the expression of light and shade of an image, and the like. These adjustments are automatically executed by various programs such as a driver program and an application program of the image reading apparatus installed in the computer apparatus, or are executed by a user or the like. In addition, there is an adjustment for correcting a backlight image. Various methods have been proposed as the backlight correction method (see Patent Documents 1 to 3).
Japanese Patent Laid-Open No. 10-79885 JP 2004-56416 A JP 2000-134467 A

  However, in the conventional method of correcting a backlight image, sufficient correction processing cannot be executed for an image whose color balance is lost, such as a halftone. That is, with the conventional correction methods, image quality cannot be improved sufficiently for such an image whose color balance is lost such as halftone. For this reason, an appropriate correction process for an image in which the color balance such as halftone is lost has been desired.

  The present invention has been made in view of such circumstances, and an object of the present invention is to perform an appropriate correction process on an image whose color balance is lost, such as a halftone.

The main invention for achieving the object is as follows:
(A) a histogram data generation unit that generates histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixels based on data of pixels constituting the image;
(B) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small region from the three or more small regions and acquires attribute information about the small region for each color;
(C) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small region, based on the attribute information acquired by the attribute information acquisition unit;
(D) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small area becomes a target value common to the colors acquired by the target value acquisition unit;
An image correction apparatus including (E).

  Other features of the present invention will become apparent from the description of the present specification and the accompanying drawings.

  At least the following matters will become apparent from the description of the present specification and the accompanying drawings.

(A) a histogram data generation unit that generates histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixels based on data of pixels constituting the image;
(B) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small region from the three or more small regions and acquires attribute information about the small region for each color;
(C) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small region, based on the attribute information acquired by the attribute information acquisition unit;
(D) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small area becomes a target value common to the colors acquired by the target value acquisition unit;
An image correction apparatus comprising (E).

  In such an image correction apparatus, based on the histogram data for each color generated based on the data of the pixels constituting the image, three regions represented by the histogram for each color are classified into three areas according to the size of the density value. The attribute information regarding at least one small area selected from these three or more small areas is obtained for each color by dividing into the above small areas, and three or more based on the acquired attribute information for each color. Attribute information on selected small areas selected from each small area by selecting at least one small area having a density value larger than a small area and smaller density value than other small areas as a selected small area By obtaining a target value common to each color and performing correction processing on the image so that the attribute information of each color related to the selected small area becomes the target value common to each color, colors such as halftones Can be subjected to appropriate correction processing on the image lance collapsed.

  In such an image correction apparatus, histogram data for each color of red, green, and blue may be generated by the histogram data generation unit as the histogram data for each color. As described above, if histogram data for each color of red, green and blue is generated as histogram data for each color, it has density value data for each color of red, green and blue as pixel data constituting the image. In this case, an appropriate correction process can be performed on an image in which the color balance such as halftone is lost.

  In such an image correction apparatus, the three or more small regions may include two or more small regions having the same area. In this way, if two or more small regions having the same area are included in three or more small regions, more appropriate correction processing is performed on an image in which the color balance such as halftone is lost. be able to.

  In such an image correction apparatus, the three or more small regions may include two or more small regions having the same number of pixels. As described above, if two or more small areas having the same number of pixels are included in three or more small areas, more appropriate correction processing is performed on an image in which the color balance such as halftone is lost. Can be applied.

  In the image correction apparatus, the three or more small regions include two or more small regions in which at least one of the number and area of the pixels is equal to each other, and the attribute The information acquisition unit selects at least one small region from two or more small regions in which at least one of the number and area of the pixels is equal to each other, and acquires the attribute information regarding the small region, The target value acquisition unit may select the selected small region from two or more small regions in which at least one of the number and area of the pixels is equal to each other. As described above, the three or more small regions include two or more small regions in which at least one of the number and area of the pixels is equal to each other, and the attribute information acquisition unit includes the number of pixels and the area. And selecting at least one small region from two or more small regions in which at least one of them is equal to each other to acquire attribute information related to the small region, and the target value acquisition unit includes the number of pixels and the area If a selected small region is selected from two or more small regions in which at least one of them is equal to each other, more appropriate correction processing can be performed on an image in which color balance such as halftone is lost.

  In the image correction apparatus, the selected small area may be located in a halftone area of the histogram. Since the selected small area is located in the halftone area of the histogram in this way, more appropriate correction processing can be performed on an image in which the color balance such as halftone is lost.

  In the image correction apparatus, the attribute information acquisition unit may acquire an average value of the density values for each of the at least one small region as the attribute information. In this way, if the attribute information acquisition unit acquires, as attribute information, an average value of density values for at least one small region for each color, more appropriate correction processing for an image in which the color balance is lost, such as a halftone Can be applied.

  Further, in the image correction apparatus, the target value acquisition unit is information on luminance for each of the at least one small region based on an average value of the density values for each color acquired by the attribute information acquisition unit. And the target value may be acquired based on the information on the luminance. In this way, the target value acquisition unit acquires information on luminance for at least one small region based on the average value of the density values for each color acquired by the attribute information acquisition unit, and the target value is acquired based on the information on the luminance. If the value is acquired, more appropriate correction processing can be performed on an image in which the color balance such as halftone is lost.

  In the image correction apparatus, the target value acquisition unit may acquire an average value of pixel luminances for the at least one small region as information on the luminance. In this way, if the target value acquisition unit acquires the average value of the luminance of the pixels for at least one small area as information relating to the luminance, more appropriate correction processing is performed on an image in which the color balance such as halftone is lost. Can be applied.

  In the image correction apparatus, the target value acquisition unit includes at least one of the average luminance values acquired for the at least one small region and the average luminance value of the entire image. The target value may be acquired based on a value obtained by adding. In this way, the target value acquisition unit is based on a value obtained by adding at least one of the average luminance values acquired for at least one small region and the average luminance value of the entire image, If the target value is acquired, more appropriate correction processing can be performed on an image in which color balance such as halftone is lost.

  In the image correction apparatus, the correction processing unit may convert the density value of each pixel constituting the image when the correction process is performed on the image. As described above, when the correction processing unit performs the correction process on the image, the correction process can be easily performed by converting the density value of each pixel constituting the image.

  Further, in such an image correction apparatus, the correction processing unit obtains information on a correspondence relationship between the density value before conversion and the density value after conversion in order to perform the correction process on the image. You may do it. If such information is acquired by the correction processing unit, the correction process can be easily performed.

  The image correction apparatus may further include a determination unit that determines whether or not the image is an image in which color balance in the selected small area is lost. If such a determination unit is provided, it is possible to determine whether or not the image is an image in which the color balance in the selected small area is lost.

  Further, in such an image correction apparatus, the correction processing unit, with respect to the image, when the determination unit determines that the image is an image in which color balance in the selected small area is lost. Correction processing may be performed. As described above, when the determination unit determines that the image is an image in which the color balance in the selected small area is lost, if the correction process is performed on the image, the image in which the color balance in the selected small area is lost is corrected. Appropriate correction processing can be performed.

(A) a histogram data generation unit that generates histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixels based on data of pixels constituting the image;
(B) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small region from the three or more small regions and acquires attribute information about the small region for each color;
(C) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small region, based on the attribute information acquired by the attribute information acquisition unit;
(D) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small area becomes a target value common to the colors acquired by the target value acquisition unit;
(E)
(F) As the histogram data for each color, histogram data for each color of red, green, and blue is generated by the histogram data generation unit,
(G) The three or more subregions include two or more subregions having the same area.
(H) The three or more subregions include two or more subregions having the same number of pixels.
(I) The three or more small regions include two or more small regions in which at least one of the number and area of the pixels is equal to each other, and the attribute information acquisition unit Selecting at least one small region from two or more small regions having at least one of a number and an area equal to each other, and acquiring the attribute information related to the small region; Selecting the selected subregion from two or more subregions in which at least one of the number and area of pixels is equal to each other;
(J) The attribute information acquisition unit acquires, as the attribute information, an average value of the density values for each color for each of the at least one small region,
(K) the selected small area is located in a halftone area of the histogram;
(L) The attribute information acquisition unit acquires, as the attribute information, an average value of the density values for each color for each of the at least one small region,
(M) The target value acquisition unit acquires information on luminance for each of the at least one small region based on an average value of the density values for each color acquired by the attribute information acquisition unit, and relates to the luminance Obtaining the target value based on the information,
(N) The target value acquisition unit acquires an average value of luminance of pixels for the at least one small region as information on the luminance,
(O) The target value acquisition unit is obtained by adding at least one of the average luminance values acquired for the at least one small region and the average luminance value of the entire image. Based on the value, obtain the target value,
(P) When performing the correction process on the image, the correction processing unit converts a density value of each pixel constituting the image,
(Q) In order to perform the correction process on the image, the correction processing unit obtains information regarding a correspondence relationship between a density value before conversion and a density value after conversion,
(R) a determination unit for determining whether or not the image is an image in which the color balance in the selected small area is lost,
(S) The correction processing unit performs the correction processing on the image when the determination unit determines that the color balance of the selected small region is lost for the image. An image correction device.

(A) generating histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel based on the data of the pixels constituting the image;
(B) Based on the generated histogram data for each color, dividing the region represented by the histogram for each color into three or more small regions for each density value;
(C) selecting at least one small area from among the three or more small areas and acquiring attribute information relating to the small area for each color;
(D) From among the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. Steps,
(E) acquiring a common target value for each color for the attribute information related to the selected small area based on the acquired attribute information;
(F) performing a correction process on the image so that the attribute information of each color related to the selected small area becomes a target value common to the acquired colors;
An image correction method comprising (G).

(A) An image correction program executed by the image correction apparatus,
(B) generating histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel based on the data of the pixels constituting the image;
(C) dividing the region represented by the histogram for each color into three or more small regions according to the density values based on the generated histogram data for each color;
(D) selecting at least one small area from the three or more small areas, and obtaining attribute information about the small area for each color;
(E) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. Steps,
(F) acquiring a common target value for each color for the attribute information related to the selected small area based on the acquired attribute information;
(G) performing a correction process on the image so that the attribute information of each color relating to the selected small area becomes a target value common to the acquired colors;
An image correction program that executes (H).

(A) an image reading unit for reading an image from a document;
(B) Histogram data generation that generates histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel based on the data of the pixels constituting the image read by the image reading unit And
(C) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small area from the three or more small areas and acquires attribute information about the small area for each color;
(D) From among the three or more small areas, at least one small area having a density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small area, based on the attribute information acquired by the attribute information acquisition unit;
(E) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small region becomes a target value common to the colors acquired by the target value acquisition unit;
An image reading apparatus comprising (F).

(A) a step of reading an image from a document;
(B) generating histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel based on the data of the pixels constituting the read image;
(C) dividing the region represented by the histogram for each color into three or more small regions according to the density values based on the generated histogram data for each color;
(D) selecting at least one small area from the three or more small areas, and obtaining attribute information about the small area for each color;
(E) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. Steps,
(F) acquiring a common target value for each color for the attribute information related to the selected small area based on the acquired attribute information;
(G) performing a correction process on the image so that the attribute information of each color relating to the selected small area becomes a target value common to the acquired colors;
An image reading method comprising (H).

(A) a printing unit that prints an image on a medium;
(B) Before the image is printed by the printing unit, based on the data of the pixels constituting the image, histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel is obtained. A histogram data generation unit to generate,
(C) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small area from the three or more small areas and acquires attribute information about the small area for each color;
(D) From among the three or more small areas, at least one small area having a density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small area, based on the attribute information acquired by the attribute information acquisition unit;
(E) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small region becomes a target value common to the colors acquired by the target value acquisition unit;
A printing apparatus comprising (F).

=== Overview of Image Reading System etc. ===
The case where the image determination device according to the present invention is applied to an image reading system will be described below as an example. 1 to 4 illustrate an embodiment of an image reading system to which an image determination device is applied. FIG. 1 illustrates an embodiment of an image reading system. FIG. 2 is a diagram illustrating an example of the internal configuration of the image reading apparatus. FIG. 3 is a diagram illustrating an example of a system configuration of the image reading apparatus. FIG. 4 is a diagram illustrating the system configuration of the computer apparatus.

  As shown in FIG. 1, the image reading system 2 includes an image reading device 10 and a computer device 20 connected to the image reading device 10 so as to be communicable by wire or wireless. As shown in FIG. 1, the image reading device 10 is a device generally called an image scanner, and includes a document table 12 and a document table cover 14 that opens and closes the upper surface of the document table 12. A document 15 from which an image is read is set on the document table 12. The document table cover 14 is provided at the rear end portion of the document table 12 through a hinge portion 18 so as to be freely opened and closed.

  On the other hand, as shown in FIG. 1, for example, the computer device 20 includes a computer main body 22, a display device 24, and an input device 26. The computer main body 22 includes various computers such as a personal computer. Here, the computer main body 22 includes a reading device 32 such as an FD drive device 28 or a CD-ROM drive device 30 therein. In addition, the computer main body 22 may include, for example, an MO (Magnet Optical) disk drive device or a DVD drive device. The display device 24 includes various display devices such as a CRT display, a plasma display, and a liquid crystal display. The input device 26 includes a keyboard 34, a mouse 36, and the like.

<Image reading device>
As shown in FIG. 2, inside the document table 12 of the image reading apparatus 10, a carriage 40 and the carriage 40 are parallel to each other along the direction of arrow A in FIG. A driving mechanism 42 for moving the guide 40 and a guide 44 for guiding the movement while supporting the carriage 40 are provided.

  The carriage 40 has an exposure lamp 46 as a light source for irradiating light to the document 15 via the document table 12, a lens 48 on which reflected light reflected by the document 15 is incident, and the inside of the carriage 40 through the lens 48. And an image sensor 50 for receiving the reflected light taken in. The image sensor 50 includes a linear CCD sensor or the like in which photoelectric conversion elements such as photodiodes that convert optical signals into electric signals are arranged in a line. The image data read by the image sensor 50 is output to the control unit 52.

  The drive mechanism 42 includes a timing belt 54 connected to the carriage 40, a pair of pulleys 55, 56 around which the timing belt 54 is stretched, and a drive motor 58 that rotationally drives one pulley 55. Yes. The drive motor 58 is driven and controlled by a control signal from the control unit 52.

  As shown in FIG. 3, the control unit 52 includes a controller 60, a motor control unit 62, a lamp control unit 64, a sensor control unit 66, an AFE (Analog Front End) unit 68, a digital processing circuit 70, And an interface circuit 72. Further, the AFE (Analog Front End) unit 68 includes an analog signal processing circuit 74 and an A / D conversion circuit 76.

  The controller 60 controls the motor control unit 62, the lamp control unit 64, the sensor control unit 66, the AFE (Analog Front End) unit 68, the digital processing circuit 70, and the interface circuit 72 based on commands from the computer main body 22. The motor control unit 62 performs drive control of the drive motor 58 for moving the carriage 40 according to a command from the controller 60. The lamp control unit 64 controls the light emission of the exposure lamp 46. The sensor control unit 66 controls the image sensor 50.

  An analog signal processing circuit 74 of an AFE (Analog Front End) unit 68 performs signal processing on an analog signal of an image read by the image sensor 50. An A / D conversion circuit 76 of an AFE (Analog Front End) unit 68 performs A / D conversion on the image signal processed by the analog signal processing circuit 74 into a digital signal.

  The digital processing circuit 70 performs digital signal processing on the digital signal sent from the A / D conversion circuit 76 of the AFE (Analog Front End) unit 68. Here, specifically, various image processing and the like are performed, including correction processing such as shading correction. The digital signal subjected to the digital signal processing is output as image data (image data) read from the document 15 to the outside, that is, to the computer main body 22 to which the image reading apparatus 10 is connected in this case. The In addition to this, the interface circuit 72 receives commands (commands) and the like from the computer main body 22 to the image reading apparatus 10.

<Computer body>
As shown in FIG. 4, the computer main body 22 includes a CPU 80, a memory 82, an HDD (Hard Disk Drive Device) 84, an operation input unit 86, a display control unit 88, an external communication unit 90, and a bus 92. I have. In addition, the computer main body 22 includes the CD-ROM drive device 30 and the FD drive device 28 described above. The CPU 80, the memory 82, the HDD (Hard Disk Drive Device) 84, the CD-ROM drive device 30, the FD drive device 28, the operation input unit 86, the display control unit 88, and the external communication unit 90 are buses. 92 to be communicable with each other.

  The CPU 80 performs overall control of the computer main body 22. The memory 82 is for storing various data such as a program executed by the CPU 80 and work data used by the program. An HDD (Hard Disk Drive Device) 84 stores an operating system (OS) executed by the CPU 80, various programs such as various application programs and drivers, and various other data such as image data. Yes. The operation input unit 86 is connected to the input device 26 such as the keyboard 34 and the mouse 36, and acquires information input by the user through the input device 26. Further, the display control unit 88 controls an image displayed on the screen of the display device 24 based on a command from the CPU 80. The external communication unit 90 communicates with various peripheral devices such as the image reading apparatus 10 connected to the outside of the computer main body 22.

  The CPU 80 reads out a program from an HDD (Hard Disk Drive Device) 84 and executes various programs under an operating system (OS). The programs executed here include various applications programs and various drivers for controlling the image reading device 10, the operation input unit 86, the display control unit 88, and the like.

  A driver for controlling the image reading apparatus 10 is generally called a scanner driver. The scanner driver is a program installed in the computer main body 22 through various communication lines such as the Internet and various storage media such as a CD-ROM and a floppy disk (FD). When the scanner driver is installed in the computer main body 22, the computer main body 22 functions as a control device that controls the image reading apparatus 10.

=== Scanner Driver ===
Next, an example of the user interface of the scanner driver will be described. FIG. 5 shows the main dialog box 100 of this user interface. This user interface is displayed on the display screen of the display device 24 by the CPU 80 of the computer main body 22 through the display control unit 88. The user can make various settings of the scanner driver through the input device 26 such as the keyboard 34 and the mouse 36 while looking at the dialog box 100 of the user interface displayed on the display screen of the display device 24.

  The main dialog box 100 is provided with a “mode selection field” 102, a “setting storage field” 104, an “original setting field” 106, an “output setting field” 108, and an “adjustment field” 110. ing. In the “mode selection field” 102, the user can select one mode from a plurality of modes. Here, “professional mode” is selected. In the “setting storage field” 104, the user can save or delete the current setting by clicking the “save button” or “delete button”.

  Further, in the “original setting field” 106, the user can set each of “original type” 112, “reading device” 114, and “automatic exposure” 116. In the “original type” 112, the type of original set can be selected. For example, “reflection original”, “film”, or the like can be selected. In the “reading device” 114, for example, “original table” can be selected. In the “automatic exposure” 116, an exposure setting suitable for the type of document to be read can be performed. For example, “photo orientation”, “document orientation”, and the like can be selected.

  In the “output setting field” 108, the user can make various settings related to image output. Specifically, in the “output setting field” 108, the “image type” 118 of the output image, “resolution” 120 at the time of reading, “original size” 122 at the time of reading, and “output size” 124 are set. Is possible. In “image type” 118, the number of colors of the read image can be selected from three types of color, gray scale, and monochrome. In “Resolution” 120, the resolution of the read image can be set. In “Original Size” 122, the size of the read image can be set.

  When the “scan button” 126 at the bottom of the dialog box 100 is clicked by the user, the scanner driver controls the external image reading device 10 based on information set by the user through the dialog box 100 to An image is read from a document set on the reading device 10. The image data thus read is sent to the computer main body. When the “preview button” 127 at the bottom of the dialog box 100 is clicked by the user, the scanner driver displays a preview window for displaying the image read by the image reading device 10 on the display screen of the display device 24.

  In addition, this scanner driver has a function of adjusting an image read by the image reading device 10. The read image is adjusted through the “adjustment column” 110 of the main dialog box 100. The “adjustment column” 110 is provided with four buttons and five check boxes for adjusting an image read by the image reading apparatus 10. The four buttons 128A, 128B, 128C, and 128D are an automatic exposure button 128A, a histogram adjustment button 128B, a density correction button 128C, and an image adjustment button 128D, respectively. The five check boxes 130A, 130B, 130C, 130D, and 130E are an unsharp mask filter check box 130A, a moire removal filter check box 130B, a fading restoration check box 130C, and a dust removal check box, respectively. 130D and a backlight correction check box 130E.

  The automatic exposure button 128A is clicked when it is desired to automatically adjust the exposure. The histogram adjustment button 128B is a button that is clicked when it is desired to adjust the contrast of an image. When the histogram adjustment button 128B is clicked, a histogram adjustment dialog box is called up. The density correction button 128C is a button that is clicked when it is desired to correct the density balance of the image. When the density correction button 128C is clicked, a dialog box for density correction is called. The image adjustment button 128D is a button that is clicked when it is desired to adjust the brightness, contrast, saturation, and color balance of the image. When this image adjustment button 128D is clicked, an image adjustment dialog box is called up.

  On the other hand, the unsharp mask filter check box 130A is a check box for instructing whether or not the unsharp mask filter can be used, and is checked when it is desired to sharpen the image. The moire removal filter check box 130B is a check box for instructing whether or not to use a filter that removes moire (mesh shading) generated in the scanning of a printed matter, and is checked when moire is conspicuous. The color fading restoration check box 130C is checked when restoring the color of the faded photograph. The dust removal check box 130D is checked to reduce dust on the film during film scanning. The backlight correction check box 130E is checked when it is desired to perform backlight correction on the read image. This backlight correction will be described in detail later.

=== Image adjustment ===
Next, histogram adjustment, density correction, and image adjustment for performing image adjustment will be described. 6 to 9 illustrate histogram adjustment, density correction, and image adjustment, respectively. FIG. 6 shows a histogram adjustment dialog box. FIG. 7 explains a specific adjustment outline of the histogram adjustment. FIG. 8 shows a density correction dialog box. FIG. 9 shows an image adjustment dialog box.

<Histogram adjustment>
In “histogram adjustment”, the brightness of the image is adjusted to improve the appearance of the read image. In the histogram adjustment dialog box 131, as shown in FIG. 6, a histogram display field 132 in which a histogram of an image to be edited is displayed, and a tone curve display in which a tone curve indicating the result of adjustment by the histogram is displayed. A column 134 and a gray balance adjustment column 136 for adjusting the gray balance for removing the color cast are provided. Here, the “histogram” represents the brightness and color distribution of the entire image, and is a graph representing the data distribution (number of pixels) from black to white of the image.

  The histogram display column 132 is provided with a channel column 138 for selecting the type (channel (color)) of the histogram to be displayed. In this channel column 138, it is possible to select from four types of RGB (red, green, blue) all colors, only R (red), only G (green), and only B (blue). If you want to adjust all RGB (red, green, blue) colors, select the top switch in this channel column 138, and a histogram of all RGB (red, green, blue) colors will be displayed on the right. If only the R (red) is to be adjusted, the second switch from the top of the channel field 138 is selected, and a histogram of only R (red) is displayed on the right side. When only G (green) is to be adjusted, when the third switch from the top of the channel field 138 is selected, a histogram of only G (green) is displayed on the right side. If it is desired to adjust only B (blue), selecting the fourth switch from the top of the channel field 138 displays a histogram of only B (blue) on the right side.

  When the displayed histogram is adjusted, the adjustment is performed using the three sliders 140A, 140B, and 140C provided below the displayed histogram. The three sliders 140A, 140B, and 140C are a slider 140A that adjusts shadows, a slider 140B that adjusts gamma, and a slider 140C that adjusts highlights. The slider 140A for adjusting the shadow is represented by a black triangle mark “▲”. The slider 140B for adjusting gamma is represented by a gray triangle mark. The slider 140C for adjusting the highlight is represented by a white triangle “Δ”. When adjustment is performed using these three sliders 140A, 140B, and 140C, the three sliders 140A, 140B, and 140C are individually moved in the left-right direction. Specifically, the slider 140A for adjusting the shadow is moved so as to be located slightly to the right of the left end of the histogram peak. Further, the slider 140C for adjusting the highlight is moved so as to be located slightly to the left of the right end of the histogram peak. The slider 140B for adjusting the gamma is adjusted so that the brightness of the intermediate portion is appropriate by moving it to the left and right between the slider 140A for adjusting the shadow and the slider 140C for adjusting the highlight. As a result, the overall light / dark balance of the image to be edited is improved, and the appearance of the image can be improved.

  In addition, the histogram display field 132 is provided with numerical value input fields 142A, 142B, and 142C for directly and directly specifying the positions of the three sliders 140A, 140B, and 140C. A shadow input value is input to the numerical value input field 142A. In addition, the gamma value is input to the numerical value input field 142B. In addition, a highlight input value is input to the numerical value input field 142C. Thereby, it is possible to easily specify the shadow input value, the highlight input value, and the gamma value by directly inputting the numerical value into each numerical value input field 142A, 142B, 142C.

  Further, dropper buttons 143A, 143B, and 143C are provided on the right of these three numerical value input fields 142A, 142B, and 142C, respectively. The dropper buttons 143A, 143B, and 143C are buttons for directly instructing points from the image to be edited displayed on the preview screen displayed separately from the dialog box for adjusting the histogram. In the three numerical value input fields 142A, 142B, and 142C, numerical values corresponding to points (pixels) designated from the image to be edited on the preview screen are directly input using the dropper buttons 143A, 143B, and 143C. .

  Further, two numerical value input fields 142D and 142E are provided below the two numerical value input fields 142A and 142C to which the shadow input value and the highlight input value are input, respectively. A shadow output value corresponding to the shadow input value is input to the left numeric input field 142D. Further, a highlight output value corresponding to the highlight input value is input to the numerical value input field 142E on the right side.

  For adjustment using these sliders 140A, 140B, and 140C and numerical value input fields 142A, 142B, 142C, 142D, and 142E, RGB (red, green, blue) all colors, R (red) only, G (green) ) And B (blue) only.

  FIG. 7 explains this histogram adjustment in detail. In this histogram adjustment, the shadow input value α01, the shadow output value α03, the highlight input value α02, the highlight output value α04, which are set through the sliders 140A, 140B, 140C or the numerical value input fields 142A, 142B, 142C, 142D, 142E, Based on the gamma value α05, a tone curve representing the correspondence between input data and output data as shown in the figure is defined. That is, the tone curve defined here includes a point T1 (also referred to as a shadow point) defined by the set shadow input value α01 and shadow output value α03, and a set highlight input value α02 and highlight output value α04. It is formed so as to pass through a point T2 (also referred to as a highlight point) defined by. Further, the tone curve is formed so as to swell toward one of the straight lines connecting the points T1 and T2 in accordance with the set gamma value α05. Based on the shadow input value α01, the shadow output value α03, the highlight input value α02, the highlight output value α04, and the gamma value α05 set in this way, a tone curve representing the correspondence between the input data and the output data is obtained. It is prescribed. The tone curve is defined for each color of R (red), G (green), and B (blue).

  The tone curve thus defined is displayed in the tone curve display field 134 as shown in FIG. In the tone curve display field 134, a tone curve corresponding to the adjustment result performed in the histogram display field 132 is displayed. Further, in the tone curve display field 134, the gradation outside the point T1 (shadow point) or the point T2 (highlight point) can be adjusted so that the tone curve can be finely adjusted. Specifically, the end curve shape change buttons 144A and 144B provided on the lower left portion and the upper right portion of the tone curve are clicked, and the desired end curve shape is selected from the displayed pull-down menu. It has become. Here, for example, the end curve shape can be selected from three types of “boost”, “normal”, and “soft”. Here, “boost” is selected when it is desired to remove unevenness by making the white background part white or making the black part black. “Normal” is selected when the highlight part and the shadow part are expressed as they are. “Soft” is selected when returning a pure white part to an original white background or returning a pure black part to an original black background.

  In the gray balance adjustment column 136, a slider 145 for adjusting the gray balance is provided. By moving the slider 145 left and right, the gray balance can be adjusted and the color cast can be removed.

<Density correction>
“Density correction” is an adjustment used when the expression of the shading of an image is partially changed. Specifically, in this “density correction”, the tone curve is adjusted to improve the appearance of the read image. In other words, by adjusting the density curve (tone curve) that changes to shadow (darkest part), midtone (halftone), and highlight (brightest part), the brightness and contrast of the entire image are balanced. Can be finished. For this purpose, in the density correction dialog box 150, as shown in FIG. 8, a tone curve display unit 152 and a type of tone curve (channel (color)) to be displayed on the tone curve display unit 152 are selected. Channel column 154 is provided.

  In this channel column 154, it is possible to select from all four colors of RGB (red, green, blue), only R (red), only G (green), and only B (blue). If you want to adjust all RGB (red, green, blue) colors, select the top switch in this channel field 154, and the tone curves for all RGB (red, green, blue) colors will be displayed on the right. . If only the R (red) is to be adjusted, the tone curve of only the R (red) is displayed on the right side when the second switch from the top of the channel field 154 is selected. If it is desired to adjust only G (green), selecting the third switch from the top of this channel field 154 displays a tone curve of only G (green) on the right side. If it is desired to adjust only B (blue), when the fourth switch from the top of the channel field 154 is selected, a tone curve of only B (blue) is displayed on the right side.

  The tone curve display unit 152 displays a tone curve having the horizontal axis as an input value and the vertical axis as an output value. When the output value is set so as not to change with respect to the input value, the tone curve becomes a straight line as indicated by a line L1 in the figure.

  When adjusting the tone curve, an arbitrary point is set on the tone curve displayed on the tone curve display unit 152, and the tone curve is adjusted while shifting this point in the vertical and horizontal directions. In the present embodiment, three points P1, P2, and P3 are arbitrarily set on the tone curve displayed on the tone curve display unit 152, and these three points P1, P2, and P3 are respectively set up, down, left, and right from the reference line L1. Move and move in each direction. As a result, a tone curve desired by the user is formed on the tone curve display portion 152. The coordinates of the three points P1, P2, and P3 can also be set through two numerical value input fields 156A and 156B provided on the left side of the tone curve display unit 152, respectively. Here, by inputting an input value in the upper numerical value input field 156A and inputting an output value in the lower numerical value input field 156B, the coordinates of each point P1, P2, and P3 can be set.

  Such adjustment of the tone curve is possible for each of four types of RGB (red, green, blue), R (red) only, G (green) only, and B (blue) only. The tone curve setting can be stored through a density correction setting storage field 158 provided at the top of the density correction dialog box 150.

<Image adjustment>
As shown in the image adjustment dialog box 160 in FIG. 9, the “image adjustment” includes (1) brightness adjustment, (2) contrast adjustment, (3) saturation adjustment, and (4). There are four types of adjustments: color balance adjustments. Further, “(4) Color balance adjustment” includes an adjustment between “cyan (C) -red (R)”, an adjustment between “magenta (M) -green (G)”, and “yellow”. There are three types of adjustments, with adjustments between (Y)-Blue (B) ".

(1) Brightness adjustment “(1) Brightness adjustment” is performed when an image is too bright or too dark. “(1) Brightness adjustment” can be performed by moving the slider 162A to the left or right, or by directly inputting a numerical value into the numerical value input field 164A provided on the right side of the slider 162A.

(2) Contrast adjustment “(2) Contrast adjustment” is performed when the contrast is made clear or, on the contrary, the difference between light and dark is reduced. "(2) Contrast adjustment" can be performed by moving the slider 162B to the left or right, or by directly inputting a numerical value into the numerical value input field 164B provided on the right side of the slider 162B.

(3) Saturation adjustment “(3) Saturation adjustment” is performed when it is desired to make the color vivid. “(3) Saturation adjustment” can be performed by moving the slider 162C to the left or right, or by directly inputting a numerical value into the numerical value input field 164C provided on the right side of the slider 162C.

(4) Color balance adjustment “(4) Color balance adjustment” is performed when the image is reddish or bluish. “(4) Color balance adjustment” is performed by moving the sliders 162D, 162E, and 162F to the left and right, and in numerical value input fields 164D, 164E, and 164F provided on the right side of the sliders 162D, 162E, and 162F, respectively. This can be done by directly entering a numerical value. As a result, the image can be adjusted to an appropriate hue. Specifically, the strength of cyan and red (R) can be adjusted by moving the adjustment slider 162D between “cyan (C) and red (R)” to the left and right. Further, the strength of magenta (M) and green (G) can be adjusted by moving the adjustment slider 162E between “magenta (M) and green (G)” to the left and right. Further, the strength of yellow (Y) and blue (B) can be adjusted by moving the adjustment slider 162F between “yellow (Y) and blue (B)” to the left and right.

  Here, “(1) Brightness adjustment” and “(4) Color balance adjustment” are all three colors of red (R), green (G), and blue (B) or shades of each color as a whole. This is a process of performing conversion to shift. In addition, “(2) Contrast adjustment” is a process for performing conversion to increase or decrease the change in shading for all three colors of red (R), green (G), and blue (B).

On the other hand, “(3) Saturation adjustment” uses the following conversion formulas (1) to (3), for example, for each color of red (R), green (G), and blue (B). This is a process of converting data. Here, input data of each color of red (R), green (G), and blue (B) is indicated by “R”, “G”, and “B”, respectively. Further, output data of each color of red (R), green (G), and blue (B) are indicated by “R ′”, “G ′”, and “B ′”, respectively.
R ′ = S11 × R + S12 × G + S13 × B (1)
G ′ = S21 × R + S22 × G + S23 × B (2)
B ′ = S31 × R + S32 × G + S33 × B (3)
Here, S11, S12, S13, S21, S22, S23, S31, S32, and S33 are coefficients set according to the set saturation value. When increasing the saturation, values larger than “1” are set in S11, S22, and S33, and negative values are set in S12, S13, S21, S23, S31, and S32. In this way, “(3) Saturation adjustment” is executed.

<Setting data>
FIG. 10A to FIG. 10C illustrate data set by the histogram adjustment, density correction, and image adjustment, respectively. FIG. 10A explains data set by histogram adjustment. FIG. 10B explains data set by density correction. FIG. 10C explains data set by image adjustment.

  In the case of histogram adjustment, as shown in FIG. 10A, the shadow input values α11, α21, and α31 and the shadow output values α13, α23, and α33 for each color of R (red), G (green), and B (blue) Highlight input values α12, α22, α32, highlight output values α14, α24, α34, and gamma values α15, α25, α35 are set as data. In addition to this, data α41, α42 relating to the “lower end shape” and “upper end shape” as the end shape of the tone curve, and the adjustment value α51 in gray balance adjustment are set as data. The scanner driver stores these data α11, α21, α31, α13, α12, α22, α32, α23, α33, α14, α24, α34, α15, α25, α35, α41, α42, α51 as setting data. These setting data α11, α21, α31, α13, α23, α33, α12, α22, α32, α14, α24, α34, α15, α25, α35, α41, α42, α51 will be described with reference to FIG. 6, for example. In addition to being set by the user through the histogram adjustment dialog box 131, the scanner driver may be automatically set by calculation or the like. The scanner driver uses the stored setting data α11, α21, α31, α13, α23, α33, α12, α22, α32, α14, α24, α34, α15, α25, α35, α41, α42, α51 based on the input image. To adjust the image.

  In the case of density correction, as shown in FIG. 10B, a plurality of points P1, P2, P3,... Set on the tone curve for each color of R (red), G (green), and B (blue). The input coordinates β11, β13, β21, β23, β31, β33,... And the output coordinates β12, β14, β22, β24, β32, β34,. The scanner driver inputs the input coordinates β11, β13, β21, β23, β31, β33... And the output coordinates β12, β14, β22, β24, β32, β34... Of these points P1, P2, P3. Is stored as setting data. These setting data β11, β13, β21, β23, β31, β33,..., Β12, β14, β22, β24, β32, β34,. In addition to being set by the user through the dialog box 150, it may be set automatically by calculation or the like by the scanner driver. The scanner driver executes density correction based on the stored setting data β11, β13, β21, β23, β31, β33..., Β12, β14, β22, β24, β32, β34.

  In the case of image adjustment, as shown in FIG. 10C, the setting value γ1 of “(1) Brightness adjustment”, the setting value γ2 of “(2) Contrast adjustment”, and “(3) Color” Setting value γ3 of “degree adjustment” and setting values γ4, γ5, and γ6 of “(4) Color balance adjustment” are set. “(4) Color balance adjustment” includes a setting value γ4 between “cyan (C) -red (R)” and a setting value γ5 between “magenta (M) -green (G)”. And a set value γ6 between “yellow (Y) -blue (B)”. The scanner driver stores these setting values γ1, γ2, γ3, γ4, γ5, and γ6 as setting data. These setting data γ1, γ2, γ3, γ4, γ5, and γ6 are set by the user through the image adjustment dialog box 160 described with reference to FIG. May be set automatically. The scanner driver performs image adjustment based on the stored setting data γ1, γ2, γ3, γ4, γ5, and γ6.

=== Adjustment procedure ===
An example of a procedure for adjusting an input image, that is, an image read by the image reading apparatus 10 based on data set by the histogram adjustment, density correction, and image adjustment will be described. FIG. 11 shows an example of this procedure.

  The scanner driver performs histogram adjustment on the input image, that is, the image read by the image reading device 10 here (S002). In this histogram adjustment, the scanner driver adjusts R (red), G (green), and B (blue) based on tone curves defined for R (red), G (green), and B (blue), respectively. Data Rin, Gin, Bin of each pixel of the input image is converted and output for each color. Here, the scanner driver is set by the user through the histogram adjustment dialog box 131 described with reference to FIG. 6 or automatically set for each color of R (red), G (green), and B (blue). Shadow input values α11, α21, α31, shadow output values α13, α23, α33, highlight input values α12, α22, α32, highlight output values α14, α24, α34, gamma values α15, α25, α35, and others The histogram adjustment is executed based on these data with reference to the data α41 and α42 of the end shape of the tone curve and the adjustment value α51 of the gray balance adjustment. Thus, the scanner driver converts the data Rin, Gin, Bin (input data) of each pixel of the input image into output data Rout1, Gout1, Bout1, and outputs the converted data.

  After performing the histogram adjustment in this way, the scanner driver proceeds to step S004, and performs image adjustment on the image data on which the histogram adjustment has been performed (S004). Here, the scanner driver performs (1) brightness adjustment, (2) contrast adjustment, and (3) color balance adjustment as image adjustment. That is, the scanner driver sets the setting value γ1 of “(1) Brightness adjustment”, the setting value γ2 of “(2) Contrast adjustment”, and the setting values γ4, γ5 of “(4) Color balance adjustment”. , And γ6 are adjusted respectively. As a result, the scanner driver converts the output data Rout1, Gout1, and Bout1 resulting from the histogram adjustment into output data Rout2, Gout2, and Bout2, and outputs them.

  Then, after performing image adjustment (excluding (3) saturation adjustment) in this way, the scanner driver proceeds to step S006, and density correction is performed on the image data subjected to image adjustment. (S006). In this density correction, the scanner driver adjusts R (red), G (green), and B (blue) based on tone curves adjusted for the respective colors R (red), G (green), and B (blue). Data of each pixel of the input image is converted and output for each color. That is, here, the scanner driver inputs the input coordinates β11 of a plurality of points P1, P2, P3,... Set on the tone curve for each color of R (red), G (green), and B (blue). Referring to the setting data of β13, β21, β23, β31, β33 ………… and output coordinates β12, β14, β22, β24, β32, β34 …………, the tone curve formed based on these setting data Based on the above, density correction is executed. As a result, the scanner driver converts the output data Rout2, Gout2, and Bout2 resulting from image adjustment (excluding (3) saturation adjustment) into output data Rout3, Gout3, and Bout3 and outputs them.

  After performing the density correction in this manner, the scanner driver proceeds to step S008, and performs “(3) Saturation adjustment” as image adjustment on the image data subjected to density correction ( S008). Here, the scanner driver performs adjustment based on the set value γ3 of “(3) Saturation adjustment”. As a result, the scanner driver converts the output data Rout3, Gout3, and Bout3 resulting from the density correction into output data Rout4, Gout4, and Bout4 and outputs them.

After performing “(3) saturation adjustment” as image adjustment in this way, the scanner driver then performs color conversion on the image data that has been subjected to “(3) saturation adjustment”. Processing is performed (S010). The color conversion process is a process for converting data into data suitable for handling by various output devices (here, display devices, various printers, etc.). Specifically, for example, the following conversion formulas (4) to (6) are used.
R '= A11 * R + A12 * G + A13 * B (4)
G ′ = A21 × R + A22 × G + A23 × B (5)
B '= A31 * R + A32 * G + A33 * B (6)
Here, input data of red (R), green (G), and blue (B) before conversion are indicated by “R”, “G”, and “B”, respectively. Further, output data of each color of red (R), green (G), and blue (B) after conversion is indicated by “R ′”, “G ′”, and “B ′”, respectively. A11, A12, A13, A21, A22, A23, A31, A32, and A33 are coefficients that are appropriately set according to the characteristics of various output devices (display device 24, printer, etc.).

  In this way, the scanner driver performs color conversion processing according to the characteristics of various output devices on the image data on which “(3) saturation adjustment” has been performed. As a result, the scanner driver converts the output data Rout4, Gout4, and Bout4 by image adjustment ((3) saturation adjustment) into output data Rout5, Gout5, and Bout5 and outputs them. Then, after performing the color conversion process in this manner, the scanner driver outputs the image subjected to the color conversion process as an output image.

  Although the case where the color conversion process is executed at the final stage has been described as an example here, the color conversion process is executed as necessary.

=== Pre-scan ===
FIG. 12 illustrates an example of an image reading procedure performed by the image reading apparatus 10. When an image is read by the image reading device 10, a pre-scan may be executed. For example, in the case of reading an image at a high resolution, this pre-scan does not execute an operation for reading an image at a high resolution from the beginning, but an operation for reading an image at a high resolution (main scan). For example, reading an image at a low resolution once.

  As shown in the figure, the pre-scan is first executed (S050). The scanner driver acquires a pre-scan image (pre-image) by this pre-scan operation (S052). Next, the scanner driver performs automatic adjustment or the like on the acquired pre-scan image (pre-image). Here, the scanner driver obtains an appropriate adjustment value in histogram adjustment, density correction, image adjustment, and the like for the acquired pre-scan image (pre-image) and automatically corrects it (S054). The automatically corrected image is displayed on the display device 24, for example.

  The user performs various adjustments (corrections) while viewing the pre-scan image (pre-image) displayed on the display device 24 or the like in this way (S056). Here, the user performs various adjustments (corrections) through the histogram adjustment dialog box 131 in FIG. 6, the density correction dialog box 150 in FIG. 8, the image adjustment dialog box 160 in FIG.

  After various adjustments (corrections) are made by the user in this way, the main scan is executed. In this main scan, the image is read from the document 15 with high resolution by the image reading device 10 (S058). The scanner driver performs various adjustments (corrections) such as histogram adjustment, density correction, and image adjustment based on the data set by the user or the like for the high-resolution image acquired by the main scan in this way. To do. Thereby, the main image subjected to various adjustments (corrections) is acquired (S060).

=== Conventional Problems and Solutions ===
<Conventional problems>
Even when various adjustments (corrections) are automatically performed on the image read by the image reading device by the histogram adjustment, density correction, image adjustment, etc., the color balance such as halftone is lost. However, sufficient correction processing could not be executed for the obtained image. Further, it has been extremely difficult for the user to perform sufficient adjustment (correction) on an image whose color balance has been lost, such as halftone, through the histogram adjustment, density correction, image adjustment, and the like.

  Therefore, it is necessary to provide a correction function for such an image so that sufficient adjustment (correction) can be performed even on an image in which color balance such as halftone is lost. there were.

<Solution>
For this reason, in the present embodiment, for example, the following correction method is performed in order to execute an appropriate correction process on an image whose color balance is lost, such as a halftone. The correction method performed here will be described in detail below.

=== Image Correction ===
A method for correcting an image in which the color balance such as halftone is lost, which is performed in the present embodiment, will be described in detail. In the present embodiment, correction processing for an image whose color balance is lost, such as halftone, is performed by a scanner driver. Therefore, the scanner driver corresponds to an “image correction program”. The computer device 20 on which the scanner driver is executed corresponds to an “image correction device”.

  FIG. 13 illustrates an image correction method performed here. The scanner driver first generates histogram data based on the image data read by the image reading device 10 (S102). From this, the scanner driver corresponds to a “histogram data generation unit”. The histogram generated here is a graph representing the distribution of the number of pixels with respect to the density value of each pixel constituting the image read by the image reading device 10. On the horizontal axis of the histogram, the density value of the pixel is set, and on the vertical axis of the histogram, the number of pixels is set. In this histogram, a rectangular bar graph representing the number of pixels for each density value on the horizontal axis of the graph is formed. The bar graphs and the like formed here are connected to each other in the horizontal direction to form a graph having an area having a certain shape as a whole.

  After generating the histogram data in this manner, the scanner driver then divides the area represented by the histogram into three or more small areas according to the density value of the pixel based on the generated histogram data. (S104). Here, the number of small areas to be divided may be four, or of course five or more. Further, the areas of the divided small regions may be set to be substantially equal to each other, or may be set to be not equal to each other. Of course, the areas of two or more of the partial small regions may be set to be substantially equal to each other. In addition, the number of pixels in the small area to be divided may be set to be substantially equal to each other, or may be set to be not equal to each other. Needless to say, the number of pixels in a part of two or more of the divided two or more small regions may be set to be substantially equal to each other.

  The scanner driver thus divides the area represented by the histogram into three or more small areas, and then selects at least one small area from the three or more divided small areas (S106). ). Here, the small area selected by the scanner driver may be two, or may be three or more. Of course, the scanner driver may select all the small areas here. Here, the small area selected by the scanner driver is necessary for the scanner driver to perform correction processing on an image read by the image reading device 10 for an image whose color balance is lost, such as halftone. A small area from which information can be extracted is desirable.

  After selecting at least one small area from the two or more small areas obtained by dividing the area represented by the histogram in this way, the scanner driver then selects attribute information from the selected small area. Is acquired (S108). Here, the scanner driver corresponds to an “attribute information acquisition unit”.

  Here, the attribute information acquired by the scanner driver is information on the properties and characteristics of the small area selected by the scanner driver. Here, as the attribute information acquired by the scanner driver, for example, the average value of the density values of the pixels occupying the small area selected by the scanner driver, or the upper limit value or the lower limit of the density value of the pixels occupying the small area selected by the scanner driver. In addition, various attribute information such as the density value corresponding to the boundary line between the small area adjacent to the small area selected by the scanner driver and the area size of the small area selected by the scanner driver. There is. The scanner driver acquires at least one type of attribute information from these various attribute information. That is, the scanner driver may acquire one type of attribute information from these various types of attribute information, or may acquire a plurality of types of attribute information.

  The scanner driver may individually acquire attribute information for each of the small areas selected from three or more small areas obtained by dividing the area represented by the histogram. Attribute information regarding two or more small regions may be acquired. In this manner, the scanner driver acquires information necessary for the scanner driver to perform correction processing for an image read by the image reading apparatus 10 for an image in which color balance such as halftone is lost.

  The scanner driver acquires the attribute information for the small area selected from the three or more small areas obtained by classifying the area represented by the histogram, and then, based on the acquired attribute information, A target value is acquired (S110). Here, the scanner driver corresponds to a “target value acquisition unit”. This target value is a target value for attribute information related to a selected small area selected from among three or more small areas obtained by classification. As the selected small region, a small region having a density value larger than a certain small region and a smaller density value than the other small regions is selected. The target value is a value for performing an appropriate correction process on an image in which the color balance in such a selected small area is lost. Here, the target value acquired by the scanner driver is a common target value for each color.

  There are various methods for acquiring the target value based on the attribute information acquired by the scanner driver. Specifically, for example, there is a method of acquiring information regarding luminance from the acquired attribute information and acquiring a target value based on the information regarding luminance. Thus, if the information regarding luminance is acquired, the target value can be easily acquired. As a method for acquiring information on the luminance, it is possible to easily acquire from the density values of the colors of red (R), green (G), and blue (B).

  After acquiring the target value in this way, the scanner driver next performs a correction process on the image based on the acquired target value common to each color (S112). Here, the scanner driver corresponds to a “correction processing unit”. The scanner driver performs correction processing so that the attribute information of each color related to the selected small area becomes the target value common to each acquired color. Various correction processes can be adopted as the correction process executed by the scanner driver. Specifically, for example, when the scanner driver performs correction processing by the above-described “density correction”, the attribute information of each color related to the selected small area is set to a target value common to each acquired color. Adjust the tone curve in “Density correction”. For this purpose, the scanner driver sets an arbitrary point for adjusting the tone curve in “density correction”. In addition, the scanner driver performs correction processing by changing parameters of various adjustments (corrections) that have already been performed on the image read by the image reading device 10.

Here, the image on which the scanner driver performs correction processing may be an image that has been subjected to various adjustments (correction) such as histogram adjustment, density correction, and image adjustment, and is read by the image reading device 10. It may be an image obtained. In this manner, the scanner driver performs a correction process for an image whose color balance in the selected small area is lost on the image read by the image reading device 10. Thereby, it is possible to improve the image quality of an image whose color balance is broken, such as a halftone. Thereafter, the scanner driver immediately ends the process.
From these facts, this scanner driver corresponds to an “image correction program”.

=== Application to actual processing ===
In the present embodiment, correction of an image whose color balance is lost, such as halftone, is performed along with various image determination processing and other image correction processing, as well as the main dialog box 100 of the user interface described with reference to FIG. This is executed when the “backlight correction check box 130E” is checked by the user.

  The scanner driver processes the image that has been adjusted (corrected) automatically or by the user by the above-described histogram adjustment, density correction, image adjustment, or the like. The scanner driver performs various determination processes on the image read by the image reading device 10, and then performs a correction process on the image according to the determination result.

Here, the determination process executed by the scanner driver is as follows.
(1) Judgment process of whether or not it is a backlight image (2) Judgment process of whether or not it is an image whose halftone color balance has been lost (3) Judgment process of whether it is an image of sunset or sunrise The correction process executed by the scanner driver is as follows.

(4) Correction processing for normal backlight image (5) Correction processing for backlight image in which halftone color balance is lost (6) Correction processing for backlight image of sunset or sunrise

  FIG. 14 is a flowchart illustrating a procedure of processing executed by the scanner driver in the present embodiment. The scanner driver first determines whether the image read by the image reading device 10 is a backlight image (S202). Here, when it is determined that the image read by the image reading device 10 is not a backlight image, the scanner driver proceeds to step S210, and the image read by the image reading device 10 is “normal”. The backlight image correction process "is executed (S210). In the present embodiment, the “normal backlight image correction process” executed here is a correction process that does not have a significant effect even when executed on an image that is not a backlight image. In other words, when the image read by the image reading device 10 is not a backlight image, even if “normal backlight image correction processing” is performed on the image, no significant adverse effect is caused. After executing the “normal backlight image correction process” on the image read by the image reading apparatus 10 in step S210, the scanner driver immediately ends the process.

  On the other hand, when the scanner driver determines in step S202 that the image read by the image reading device 10 is a backlight image, the scanner driver proceeds to step S204. Here, the scanner driver determines whether or not the image read by the image reading apparatus 10 is an image in which the halftone color balance is lost (S204). For this reason, the scanner driver corresponds to a “determination unit”. Here, when the scanner driver determines that the image read by the image reading device 10 is an image in which the halftone color balance is lost, the process proceeds to step S212, where the image reading device 10 reads the image. “Correction processing for a backlight image in which the halftone color balance is lost” is executed on the image (S212). This “correction processing for a backlight image in which the halftone color balance is lost” is different from the above-mentioned “normal backlight image correction processing”, and the image quality of the backlight image in which the halftone color balance is lost. It is executed to improve. After executing “correction processing for a backlight image in which the halftone color balance is lost” on the image read by the image reading apparatus 10 in this manner, the scanner driver immediately ends the processing.

  On the other hand, if the scanner driver determines in step S204 that the image read by the image reading device 10 is not an image in which the halftone color balance is lost, the process proceeds to step S206. Then, the scanner driver next executes a process of determining whether or not the image read by the image reading device 10 is a sunset or sunrise image (S206). Here, if the scanner driver determines that the image read by the image reading device 10 is a sunset or sunrise image, the scanner driver proceeds to step S <b> 208, and the image read by the image reading device 10 is processed. , “Correction process for backlight image of sunset or sunrise” is executed (S208). This "correction processing for sunset or sunrise backlight images" is different from the above-mentioned "normal backlight image correction processing" and "correction processing for backlight images in which the halftone color balance is lost". is there. This correction process is executed to improve the image quality of the backlight image of the sunset or sunrise. After executing the “correction process for a sunset or sunrise backlight image” on the image read by the image reading apparatus 10 in this manner, the scanner driver immediately ends the process.

  On the other hand, if the scanner driver determines in step S206 that the image read by the image reading device 10 is not a sunset or sunrise image, the scanner driver proceeds to step S210, and the image reading device 10 A “normal backlight image correction process” is executed on the read image (S210). This is because the image read by the image reading device 10 is determined to be neither an image whose halftone color balance is lost nor an image of sunset or sunrise, This is because it is determined that the image is a “normal backlight image”. For this reason, the scanner driver performs “normal backlight image correction processing” on the image read by the image reading device 10. In this way, the scanner driver performs the “normal backlight image correction process” on the image read by the image reading apparatus 10 and then immediately ends the process.

  The scanner driver according to the present embodiment performs various image determination processes and various image correction processes on the image read by the image reading apparatus 10 in the above procedure. Hereinafter, various image determination processes and various image correction processes executed by the scanner driver will be individually described in detail.

=== Determination of Backlight Image ===
When the “backlight correction check box 130E” is checked by the user in the main dialog box 100 of the user interface described with reference to FIG. For the image read in step 10, “determination processing for determining whether the image is a backlight image” is performed. The “determination process for determining whether or not the image is a backlight image” performed here will be described in detail below.

<Judgment procedure>
FIG. 15 illustrates an example of a determination procedure of “determination process for determining whether the image is a backlight image” performed by the scanner driver. Based on the image data read by the image reading device 10, histogram data is generated (S302). The histogram generated here is the same as the histogram described above. That is, this histogram is a graph representing the distribution of the number of pixels with respect to the density value of each pixel constituting the image read by the image reading device 10.

  After generating the histogram data in this manner, the scanner driver then divides the area represented by the histogram into three or more small areas according to the density value of the pixel based on the generated histogram data. (S304). Here, the number of small areas to be divided may be four, or of course five or more. Further, the areas of the divided small regions may be set to be substantially equal to each other, or may be set to be not equal to each other. Of course, some of the three or more divided small regions, that is, two or more small regions may be set to have substantially the same area. In addition, the number of pixels in the small area to be divided may be set to be substantially equal to each other, or may be set to be not equal to each other. Of course, it may be set such that the number of pixels in a part of three or more divided small regions, that is, two or more small regions, is substantially equal to each other.

  After the scanner driver divides the area represented by the histogram into three or more small areas in this way, the first small area and the second small area are then selected from the divided three or more small areas. Is selected (S306). Here, the first small region is at least one small region selected from the three or more divided small regions. This first small region is selected in order to obtain necessary information from the image read by the image reading device 10 and determine whether or not it is a backlight image. The number of small regions selected as the first small region may be one, two, or even three or more.

  The second small area is at least one small area selected from the three or more divided small areas, as in the case of the first small area. However, here, a small region having a density value larger than that of the first small region and not adjacent to the first small region is selected as the second small region. The same applies to the case where there are two or more first small areas. That is, even when there are two or more first small areas, the second small areas are not adjacent to the two or more first small areas, respectively, and the two or more first small areas are not adjacent to each other. The density value of the second small region is larger than that. The same applies when there are two or more second small regions. That is, both of these two or more second small regions are not individually adjacent to one or two or more first small regions, and the one or two or more first small regions are not adjacent to each other. On the other hand, the density value of each second small region is large. The number of small regions selected as the second small region may be one, two, or even three or more. Further, as in the case of the first small area, this second small area is selected in order to obtain necessary information from the image read by the image reading device 10 and determine whether or not it is a backlight image.

  After selecting the first small area and the second small area from among the three or more small areas in this way, the scanner driver then individually sets the attribute for each of the first small area and the second small area. Information is acquired (S308). Here, the attribute information is a property or characteristic of the first small area or the second small area. Here, the attribute information acquired by the scanner driver includes an average value of density values of pixels occupying the first small area or the second small area, and an upper limit value of density values of pixels occupying the first small area or the second small area. Alternatively, there are various attribute information such as a lower limit value, a density value corresponding to a boundary line between adjacent small areas, and the size of the area of the first small area or the second small area. The scanner driver may acquire one type of attribute information among these various types of attribute information, or may acquire a plurality of types of attribute information. The scanner driver acquires information necessary for determining whether the image read by the image reading device 10 is a backlight image as attribute information from the first small area and the second small area.

  Then, the scanner driver determines whether or not the image read by the image reading device 10 is a backlight image based on the attribute information acquired from the first small area and the second small area in this way (S310). Here, for example, the scanner driver compares the attribute information acquired from the first small area with the attribute information acquired from the second small area, and based on the comparison result, the image read by the image reading device 10 It may be determined whether the image is a backlight image. Specifically, the scanner driver obtains a difference between a value acquired as attribute information from the first small area and a value acquired as attribute information from the second small area, and based on this difference, for example, the difference is predetermined. By checking whether or not the value is exceeded, it is determined whether or not the image read by the image reading device 10 is a backlight image. As a result, the scanner driver ends the process of determining whether the image read by the image reading device 10 is a backlight image.

<Generation of histogram>
FIG. 16 illustrates an example of a histogram generated by the scanner driver. In the present embodiment, the data of each pixel constituting the image read by the image reading device 10 is constituted by data of density values of three colors of red (R), green (G), and blue (B). . For this reason, in this embodiment, a histogram is generated for each color of red (R), green (G), and blue (B). That is, an image is formed with a histogram generated based on the red (R) density value of each pixel constituting the image, and a histogram generated based on the green (G) density value of each pixel constituting the image. Three types of histograms are generated: a histogram generated based on the blue (B) density value of each pixel. The scanner driver determines whether or not the image is a backlight image based on the three-color histogram data of red (R), green (G), and blue (B) generated in this way.

<Division into small areas>
Next, the scanner driver generates three or more regions represented by the histograms of the respective colors based on the three-color histogram data of red (R), green (G), and blue (B) generated in this way. Execute processing to divide into small areas. Here, the scanner driver classifies the region represented by the histogram of each color according to the density value. As a result, as shown in FIG. 16, the three or more small regions are divided by boundary lines provided corresponding to density values along the direction of the vertical axis of the histogram of each color. Arranged side by side along the direction of the axis.

  In this embodiment, the scanner driver divides an area represented by a histogram of three colors of red (R), green (G), and blue (B) into four small areas Ry1 and Ry2 as shown in FIG. , Ry3 and Ry4. However, of the four small regions Ry1, Ry2, Ry3, and Ry4 divided here, the three small regions Ry1, Ry2, and Ry3 are set so that their areas are substantially equal to each other. That is, these three small regions Ry1, Ry2, and Ry3 have substantially the same number of pixels.

  Here, the areas of these three small regions Ry1, Ry2, and Ry3 each occupy approximately 33% of the total number of pixels constituting the image. That is, since the small region Ry1 is located on the side where the density value is the smallest, among the pixels constituting the image to be determined, pixels from “0 to 33%” in order from the pixel having the smallest density value. Is formed. In addition, since the small area Ry2 is located on the side having the second smallest density value after the small area Ry1, the pixel next to the pixel included in the small area Ry1 among the pixels constituting the image to be determined. The pixels are formed in order of “34 to 66%” pixels. In addition, since the small area Ry3 is located on the side having the third smallest density value after the small area Ry2, the pixel next to the pixel included in the small area Ry2 among the pixels constituting the image to be determined. In this order, pixels are formed from “67 to 99%”. For these reasons, the small area Ry1 is located in the shadow area. The small area Ry2 is located in the halftone area. The small area Ry3 is located in the highlight area.

  On the other hand, the small region Ry4 is set to have a different area (number of pixels) from the three small regions Ry1, Ry2, and Ry3. Here, since the small area Ry4 is located on the side having the fourth lowest density value after the small area Ry3 and located on the side having the largest density value, among the pixels constituting the image to be determined. The remaining 1% of pixels are formed.

<Selection of small areas>
After dividing the region represented by the histogram of three colors of red (R), green (G) and blue (B) in this way into four small regions Ry1, Ry2, Ry3 and Ry4, the scanner driver Next, a process of selecting the first small area and the second small area from the four divided small areas Ry1, Ry2, Ry3, and Ry4 is executed for each color. Here, the first small area is at least one small area selected from the four divided small areas Ry1, Ry2, Ry3, and Ry4, and is not adjacent to the second small area. And a small region that satisfies the condition that the density value is smaller than that of the second small region. The second small area is at least one small area selected from the four divided small areas Ry1, Ry2, Ry3, and Ry4, and is not adjacent to the first small area. In addition, the small area satisfies the condition that the density value is larger than that of the first small area.

  In this embodiment, the scanner driver selects a small area Ry1 for each color from the four small areas Ry1, Ry2, Ry3, and Ry4 as the first small area (see FIG. 16). In this embodiment, the scanner driver selects a small region Ry3 for each color from the four small regions Ry1, Ry2, Ry3, and Ry4 as the second small region (see FIG. 16). The reason why the first small area and the second small area are selected in this way is as follows. In other words, in the present embodiment, among the four small regions Ry1, Ry2, Ry3, and Ry4, the three small regions Ry1, Ry2, and Ry3 each have an area (number of pixels) that is substantially equal to the total number of pixels. It is set to 33%. That is, the three small areas Ry1, Ry2, and Ry3 are set so as to occupy approximately 1/3 of the area represented by the histogram of each color of red (R), green (G), and blue (B). Has been. When the first small region and the second small region are selected from these three small regions Ry1, Ry2, and Ry3, the small region Ry1 is selected as the first small region from the requirement that they are not adjacent to each other, and the second small region As a result, there is only an option of selecting the small region Ry3.

  In addition, when the area | region represented by the histogram of each color of red (R), green (G), and blue (B) is divided into methods other than such a method, the first small area and The second small area may be selected.

<Acquisition of attribute information>
After selecting the small area Ry1 as the first small area for each color from the four small areas Ry1, Ry2, Ry3, and Ry4 and selecting the small area Ry3 as the second small area for each color, the scanner driver Next, attribute information is acquired for each color from the small area Ry1 selected as the first small area and the small area Ry3 selected as the second small area. In the present embodiment, the upper limit value St1 of the small area Ry1 and the upper limit value St3 of the small area Ry3 are acquired as attribute information for each color. The upper limit value St1 of each color small area Ry1 is also a density value corresponding to the boundary line between the adjacent small areas Ry2. Similarly, the upper limit value St3 of the small area Ry3 of each color is also a density value corresponding to the boundary line between the adjacent small areas Ry4. Further, the upper limit value St1 of the small area Ry1 of each color is also referred to as “33% reference point”. Further, the upper limit value St3 of the small area Ry3 of each color is also referred to as “99% reference point”. The upper limit value St2 of the small area Ry2 of each color is also referred to as “66% reference point”.

  Here, the reason why the upper limit values St1 and St3 are acquired as the attribute information of the small area Ry1 and the small area Ry3 is as follows. That is, the upper limit value St1 of the small region Ry1 and the upper limit value St3 of the small region Ry3 can be accurately determined in determining whether the image is a backlight image, as compared with other attribute information. is there. That is, the upper limit values St1 and St3 of the small area Ry1 and the small area Ry3 are important determination factors in determining whether the image is a backlight image or not, as compared with other attribute information of the small areas Ry1 and Ry3. To get.

  For example, as attribute information of the small region Ry1 selected as the first small region and the small region Ry3 selected as the second small region, the average value Av1 of the density values of the pixels occupying the small region Ry1 and the pixels occupying the small region Ry3 When the average value Av3 of the density values is acquired, the difference between the average values Av1 and Av3 is much smaller than the difference between the upper limit values St1 and St3 of the small region Ry1 and the small region Ry3. For this reason, in determining whether the image read by the image reading device 10 is a backlight image, as the attribute information of the small area Ry1 and the small area Ry3, the upper limit value St1 of the small area Ry1 and the small area Ry3 It is preferable to obtain the upper limit value St3 for each color.

<Judgment>
In this way, the region represented by the histogram of three colors of red (R), green (G) and blue (B) is divided into four small regions Ry1, Ry2, Ry3, and Ry4, respectively. The upper limit value St1 of the small area Ry1 and the upper limit value St3 of the small area Ry3 are acquired for each color as attribute information of the small area Ry1 selected as the first small area and the small area Ry3 selected as the second small area from among them. After that, the scanner driver then determines whether the image read by the image reading device 10 is a backlight image based on the acquired upper limit value St1 of the small area Ry1 for each color and the upper limit value St3 of the small area Ry3 for each color. A determination process of whether or not is executed.

Here, in the present embodiment, the scanner driver first sets the upper limit value St1 and the small region Ry3 of the small region Ry1 obtained for the three color histograms of red (R), green (G), and blue (B), respectively. From the upper limit value St3, an operation for obtaining a value when converted into luminance is performed. In this calculation, an expression used for converting the density value of each color of red (R), green (G), and blue (B) into an actual luminance value is used. An example of this arithmetic expression is shown in the following expression (7).
Actual luminance value = 1/5 × R + 3/5 × G + 1/5 × B (7)
Here, “R” indicates the density value of the red (R) component of the pixel. “G” indicates the density value of the green (G) component of the pixel. “B” indicates the density value of the blue (B) component of the pixel. In the formula (7), the ratio of each color of red (R), green (G) and blue (B) is “1: 3: 1” as an example. May be set.

  Using the equation (7), the scanner driver actually calculates from the upper limit value St1 of the small area Ry1 obtained for each of the three color histograms of red (R), green (G), and blue (B). A value St10 corresponding to the luminance value of is obtained. This value St10 is a “33% reference point” corresponding to the actual luminance value. Further, the scanner driver converts the upper luminance value St3 of the small area Ry3 obtained for each of the three color histograms of red (R), green (G), and blue (B) to the actual luminance value according to Expression (7). The corresponding value St30 is acquired. The value St30 is a “99% reference point” corresponding to the actual luminance value.

  Thereafter, the scanner driver calculates the actual luminance value St10 ("33% reference point") obtained from the upper limit value St1 of the RGB small color area Ry1 and the upper limit value St3 of the RGB small color area histogram Ry3. A difference ΔSt from the obtained actual luminance value St30 (“99% reference point”) is obtained. Here, the scanner driver determines the upper limit value St1 of the RGB small color region Ry1 from the actual luminance value St30 ("99% reference point") obtained from the upper limit value St3 of the RGB small color region Ry3. An operation for subtracting the actual luminance value St10 ("33% reference point") obtained from the above is executed.

  Then, the scanner driver compares the difference ΔSt obtained in this way with a predetermined threshold value ΔStk, and if the obtained difference ΔSt is equal to or larger than the predetermined threshold value ΔStk, the image reading device The image read by 10 is determined to be a backlight image. On the other hand, the scanner driver compares the obtained difference ΔSt with a predetermined threshold value ΔStk. If the obtained difference ΔSt has not reached the predetermined threshold value ΔStk, the scanner driver reads it. The determined image is not a backlight image. Here, the predetermined threshold value ΔStk is a value that is appropriately set for determining whether or not the image is a backlight image. This predetermined threshold value ΔStk is appropriately set based on an analysis result or a rule of thumb by a simulation or the like performed in advance.

  FIG. 17A and FIG. 17B describe a method for determining whether or not the image is a backlight image by a scanner driver. FIG. 17A illustrates a case where the scanner driver determines that the image to be determined is not a backlight image. FIG. 17B illustrates a case where the scanner driver determines that the image to be determined is a backlight image.

  When the image read by the image reading device 10 is not a backlight image, the pixel distribution state with respect to the density value represented by the histogram of three colors of red (R), green (G), and blue (B) is almost uniform. become. From this, when the image read by the image reading device 10 is not a backlight image, as shown in FIG. 17A, a value St10 that is “33% reference point” and a value St20 that is “66% reference point” Then, the value St30 which becomes the “99% reference point” is gradually increased at a small increase rate with a gentle slope. Here, the value St20 serving as the “66% reference point” is obtained from the upper limit value St2 of the small region Ry2 of the three-color histogram of red (R), green (G), and blue (B) by the above formula ( This is a value when converted to an actual luminance value using 7). For this reason, the difference ΔSta between the value St10 serving as the “33% reference point” and the value St30 serving as the “99% reference point” is smaller than the predetermined threshold value ΔStk.

  On the other hand, when the image read by the image reading device 10 is a backlight image, the pixel distribution state with respect to the density value represented by the histogram of three colors of red (R), green (G), and blue (B). There is a bias. From this, when the image read by the image reading device 10 is a backlight image, as shown in FIG. 17B, the value St10 that becomes the “33% reference point” and the value St20 that becomes the “66% reference point”. And the value St30 which becomes the “99% reference point” has an unbalanced relationship. As a result, the value St10 serving as the “33% reference point” is too small, and the value St30 serving as the “99% reference point” is too large. From this, the slope becomes steep from the value St10 which becomes the “33% reference point” to the value St30 which becomes the “99% reference point”, and the increase rate becomes large. Therefore, the difference ΔStb between the value St10 that becomes the “33% reference point” and the value St30 that becomes the “99% reference point” is a value larger than the predetermined threshold value ΔStk.

  By determining whether or not the image to be determined by the scanner driver in this way, that is, the image read by the image reading device 10 here is a backlight image, it is more accurately determined whether or not it is a backlight image. Can be determined. Specifically, for example, a general backlight image, a backlight image in which a color balance such as a halftone is lost, or a backlight image with a sunset or sunrise as a background is determined to be a smooth backlight image. be able to.

=== Judgment of an Image with Discolored Halftone Color Balance ===
Next, “determination of whether or not an image has a halftone color balance collapsed” will be described. The “determination of whether or not the image has a halftone color balance collapsed” is divided into the RGB color histograms using the RGB color histograms generated in the previous “backlight image determination”. Two or more subregions are selected from the obtained four subregions Ry1, Ry2, Ry3, and Ry4, and attribute information is acquired for each of the selected subregions. “Determining whether the image is out of color balance”. The method will be described in detail below.

<Characteristics of an image with an unbalanced halftone color balance>
FIG. 18A to FIG. 18C show examples of histograms for each of the RGB colors of an image in which the halftone color balance is lost. FIG. 18A shows a red (R) histogram. FIG. 18B shows a green (G) histogram. FIG. 18C shows a histogram of blue (B).

  In an image in which the halftone color balance is lost, any one of red (R), green (G), and blue (B) has halftone characteristics different from those of other colors. Specifically, for example, when the blue (B) halftone characteristics are different, as shown in FIGS. 18A to 18C, the blue (B) halftone region is replaced with another red (R). ) And green (G) halftone areas. Accordingly, the attributes of the blue (B) halftone area, that is, the upper limit value, the lower limit value, the average value, and the like are different from the attributes of the other red (R) and green (G) halftone areas. It will be. From this, it is possible to easily check whether or not an image has a halftone color balance broken by examining the halftone attributes of the histogram of each RGB color.

<Selection of small areas>
FIG. 19 illustrates the selection of a small area in “determination of whether or not an image has a halftone color balance collapsed”. Here, as shown in the figure, three subregions Ry1, Ry2, Ry3 out of four subregions Ry1, Ry2, Ry3, and Ry4 obtained by dividing the histogram generated in “backlight image determination” are shown. A first small area and a second small area are selected from the above. The first small region has a density value larger than at least one of the three small regions Ry1, Ry2, and Ry3, and at least one of the other three small regions Ry1, Ry2, and Ry3. This is a small area having a density value smaller than that of the small area. In the present embodiment, the small area Ry2 is selected as the first small area.

  On the other hand, the second small area is a small area selected from other small areas excluding the first small area. In the present embodiment, the small area Ry3 is selected as the second small area. Here, the reason why the small area Ry3 is selected as the second small area is that the attribute information may be more reliable than the small area Ry1. For example, when an image is read from a negative film or the like, the reliability as attribute information of the shadow area, that is, the small area Ry1 may be lowered. If the small area Ry3 is selected as the second small area, highly reliable attribute information can be acquired regardless of the type of the read image. In addition to the small region Ry3, the small region Ry1 may be selected as the second small region.

  Such selection of the first small area and the second small area is performed for each color in the histogram of each color of red (R), green (G), and blue (B).

<Acquisition of attribute information>
In this manner, the scanner driver selects the first small area and the second small area in the histogram of each color of red (R), green (G), and blue (B), and then selects the first small area. Attribute information is acquired for each color for the small area and the second small area. In this embodiment, the scanner driver uses the upper limit value St2 (also referred to as “66% reference point”) of the first small area (small area Ry2) and the upper limit value of the second small area (small area Ry3) as attribute information. St3 (also referred to as “99% reference point”) is acquired for each color.

  Here, the upper limit values St2 and St3 of the first small area (small area Ry2) and the second small area (small area Ry3) are acquired as attribute information, and the upper limit values of the first small area (small area Ry2) are acquired. When determining whether or not the upper limit value St3 of St2 and the second small area (small area Ry3) is an image in which the halftone color balance is lost compared to other attribute information, an accurate determination can be made. Because it can.

  The upper limit values of the first small area (small area Ry2) and the second small area (small area Ry3) in the red (R) histogram are “Str2” and “Str3”, respectively. Further, the upper limit values of the first small region (small region Ry2) and the second small region (small region Ry3) in the green (G) histogram are set to “Stg2” and “Stg3”, respectively. In addition, the upper limit values of the first small area (small area Ry2) and the second small area (small area Ry3) in the histogram of blue (B) are “Stb2” and “Stb3”, respectively.

<Judgment>
In this manner, the scanner driver acquires attribute information for each color from the first small area (small area Ry2) and the second small area (small area Ry3), and then, based on the acquired attribute information, the image reading device. It is determined whether the image read in step 10 is an image in which the halftone color balance is lost. Specifically, the determination is performed as follows.

  20A and 20B illustrate a method for determining whether or not an image has a halftone color balance that has been lost by the scanner driver. FIG. 20A illustrates a case where it is determined that the image has a halftone color balance collapsed. FIG. 20B illustrates a case where it is determined that the image is not an image whose halftone color balance is lost.

  Here, the case where the relationship between the upper limit values “Str2”, “Stg2”, and “Stb2” of each color of the first small region (small region Ry2) is described as an example is Str2> Stg2> Stb2. The relationship between the upper limit values “Str3”, “Stg3”, and “Stb3” of each color in the second small area (small area Ry3) will be described by taking as an example the case where Str3> Stg3> Stb3.

  In the case of an image in which the color balance of the halftone is lost, as shown in FIG. 20A, the upper limit values “Str2” and “Stg2” of each color of the halftone, that is, the first small area (small area Ry2) here. , The variation of “Stb2” is much larger than the variation of the upper limit values “Str3”, “Stg3”, and “Stb3” of each color of the second small region (small region Ry3). From this, the maximum difference ΔHs1 between the upper limit values “Str2”, “Stg2”, “Stb2” of each color of the first small area (small area Ry2) and each color of the second small area (small area Ry3) The maximum difference ΔHs2 between “Str3”, “Stg3”, and “Stb3” is obtained, and the two differences ΔHs1 and ΔHs2 are compared, whereby the halftone color balance is lost. It can be easily determined whether or not.

  Here, since the relationship of the upper limit values “Str2”, “Stg2”, “Stb2” of each color of the first small region (small region Ry2) is Str2> Stg2> Stb2, the first small region (small region Ry2) ), The maximum difference ΔHs1 between the upper limit values “Str2”, “Stg2”, and “Stb2” of each color is between the upper limit value “Str2” of red (R) and the upper limit value “Stb2” of blue (B). It can be obtained from the difference. Similarly, the maximum difference ΔHs2 between the upper limit values “Str3”, “Stg3”, and “Stb3” of each color of the second small area (small area Ry3) is similarly set to each color of the second small area (small area Ry3). Since the relationship between the upper limit values “Str3”, “Stg3”, and “Stb3” of the table is Str3> Stg3> Stb3, the upper limit value “Str3” of red (R) and the upper limit value “Stb3” of blue (B) It can be obtained from the difference between.

  On the other hand, if the image is not an image in which the color balance of the halftone is lost, as shown in FIG. The variation of “Stg2” and “Stb2” does not become very large compared to the variation of the upper limit values “Str3”, “Stg3”, and “Stb3” of each color of the second small region (small region Ry3). Therefore, the maximum difference ΔHs1 between the upper limit values “Str2”, “Stg2”, “Stb2” of each color in the first small area (small area Ry2) and the upper limit of each color in the second small area (small area Ry3) The difference from the maximum difference ΔHs2 between the values “Str3”, “Stg3”, and “Stb3” does not become so large. From this, it can be easily determined that the image is not an image in which the halftone color balance is lost.

The maximum difference ΔHs1 between the colors “Str2”, “Stg2”, “Stb2” of each color in the first small area (small area Ry2) and the upper limit value “Str3 of each color in the second small area (small area Ry3)” ”,“ Stg3 ”, and“ Stb3 ”are compared with the maximum difference ΔHs2, for example, a difference between the difference ΔHs1 and the difference ΔHs2, that is, a value obtained by subtracting the difference ΔHs2 from the difference ΔHs1 is predetermined. This can be easily done by checking whether or not the threshold value is “Ks” or more. That is, by checking whether or not the following relational expression (8) is satisfied, it is possible to easily determine whether or not the image has an intermediate color balance.
ΔHs1-ΔHs2 ≧ Ks (8)
By determining by the scanner driver by such a method, it is accurately determined whether or not the image to be determined, that is, the image read by the image reading apparatus 10 is an image in which the halftone color balance is lost. Can be determined.

<Process flow>
FIG. 21 explains the flow of determination processing by the scanner driver. As shown in the figure, the scanner driver first sets the upper limit of each color of the first small area (small area Ry2) as attribute information regarding the first small area (small area Ry2) and the second small area (small area Ry3). The values “Str2”, “Stg2”, “Stb2” and the upper limit values “Str3”, “Stg3”, “Stb3” of each color of the second small area (small area Ry3) are acquired (S402). Next, based on the obtained upper limit values “Str2”, “Stg2”, “Stb2”, “Str3”, “Stg3”, “Stb3”, the scanner driver determines each color of the first small area (small area Ry2). Between the maximum difference ΔHs1 between the upper limit values “Str2”, “Stg2”, “Stb2” and the upper limit values “Str3”, “Stg3”, “Stb3” of each color of the second small area (small area Ry3) The maximum difference ΔHs2 is obtained (S404). Next, the scanner driver compares the obtained two differences ΔHs1 and ΔHs2 (S406). Then, as a result of the comparison, it is checked whether or not the difference between these two differences ΔHs1 and ΔHs2, that is, ΔHs1−ΔHs2 is equal to or greater than a predetermined threshold “Ks” (S408). If ΔHs1−ΔHs2 is equal to or greater than the predetermined threshold value “Ks”, the scanner driver determines that the image is a halftone color balance disrupted (S410). On the other hand, if ΔHs1−ΔHs2 is not equal to or greater than the predetermined threshold “Ks”, the scanner driver determines that the image is not an image whose halftone color balance has been lost (S412). In this way, after determining whether or not the image has an intermediate color balance, the scanner driver immediately ends the process.

=== Determination of sunset / morning image ===
Next, “determination of whether the image is a sunset or sunrise image” will be described. As for “determining whether an image is a sunset or sunrise image”, as in the case of “determining whether an image is out of halftone color balance” Using the generated RGB color histograms, two or more small regions were selected from the four small regions Ry1, Ry2, Ry3, and Ry4 obtained by segmentation in the RGB color histograms. Attribute information is acquired for each of the small areas, and “determination of whether the image is a sunset or sunrise image” is performed based on the attribute information. The method will be described in detail below.

<Characteristics of sunset or sunrise image>
FIG. 22A to FIG. 22C show examples of histograms of RGB colors of sunset or sunrise images, respectively. FIG. 22A shows an example of a red (R) histogram. FIG. 22B shows an example of a green (G) histogram. FIG. 22C shows an example of a blue (B) histogram.

  The image of sunset or sunrise is different from the other colors in highlight characteristics of any one of red (R), green (G), and blue (B). In particular, for example, in the case of a sunset image, the characteristics of the red (R) highlight region are often different from those of other colors. In the case of an image of sunrise, the characteristics of blue (B) highlight are often different from those of other colors. Specifically, for example, in the case of a sunset image, as shown in FIGS. 22A to 22C, the density value of the highlighted region of red (R) is other green (G) or blue (B). It is significantly larger than Accordingly, the attributes of the red (R) highlight area, that is, for example, the upper limit value, the lower limit value, and the average value are different from the attributes of the other blue (B) and green (G) highlight areas. It will be. Similarly, for the sunrise image, for example, the attribute of the highlight area of blue (B) is different from the attribute of the highlight area of other red (R) and green (G). From this, it is possible to easily check whether the image is a sunset or sunrise image by checking the attribute of the highlight area of the histogram of each RGB color.

<Selection of small areas>
FIG. 23 illustrates the selection of a small area in “determination of whether an image is sunset or sunrise”. Here, as shown in the figure, three subregions Ry1, Ry2, Ry3 out of four subregions Ry1, Ry2, Ry3, and Ry4 obtained by dividing the histogram generated in “backlight image determination” are shown. A first small area and a second small area are selected from the above. As described above, the first small area is a small area having a density value larger than that of at least one other small area. In the present embodiment, the small area Ry3 is selected as the first small area. On the other hand, the second small region is a small region having a density value smaller than that of the first small region, as described above. In the present embodiment, the small region Ry2 is selected as the second small region. In addition to the small region Ry2, the small region Ry1 may be selected as the second small region. Further, such selection of the first small region and the second small region is performed for each color in the histogram of each color of red (R), green (G), and blue (B).

<Attribute information>
In this manner, the scanner driver selects the first small area and the second small area in the histogram of each color of red (R), green (G), and blue (B), and then selects the first small area. Attribute information is acquired for each color for the small area and the second small area. In the present embodiment, the scanner driver uses the upper limit value St3 (also referred to as “99% reference point”) of the first small area (small area Ry3) and the upper limit value of the second small area (small area Ry2) as attribute information. St 2 (also referred to as “66% reference point”) is acquired for each color. The upper limit values of the first small area (small area Ry3) and the second small area (small area Ry2) in the red (R) histogram are “Str3” and “Str2”, respectively. Further, the upper limit values of the first small area (small area Ry3) and the second small area (small area Ry2) in the green (G) histogram are set to “Stg3” and “Stg2”, respectively. The upper limit values of the first small area (small area Ry3) and the second small area (small area Ry2) in the blue (B) histogram are set to “Stb3” and “Stb2”, respectively.

<Judgment>
In this way, the scanner driver obtains attribute information for each color from the first small area (small area Ry3) and the second small area (small area Ry2), and then based on the acquired attribute information, the image reading device It is determined whether or not the image read at 10 is an image of sunset or sunrise. Specifically, the determination is performed as follows.

  FIG. 24A and FIG. 24B explain a method for determining whether the image is a sunset or sunrise image by the scanner driver. FIG. 24A illustrates a case where it is determined that the image is sunset or sunrise. FIG. 24B illustrates the case where it is determined that the image is not a sunset or sunrise image.

  Here, the case where the relationship between the upper limit values “Str3”, “Stg3”, and “Stb3” of each color of the first small region (small region Ry3) is described as an example is Str3> Stb3> Stg3. The relationship between the upper limit values “Str2”, “Stg2”, and “Stb2” of each color in the second small area (small area Ry2) will be described by taking as an example the case where Str2> Stb2> Stg2.

  In the case of an image of sunset or sunrise, as shown in FIG. 24A, highlights, that is, here, upper limit values “Str3”, “Stg3”, “Stb3” of each color of the first small region (small region Ry3). Among these, the upper limit value of a specific color is much larger than the upper limit values of other colors. That is, in the case of a sunset image, for example, the upper limit value “Str3” of the first small area (small area Ry3) of red (R) is the upper limit value of other colors, that is, blue (B) or It is much larger than the upper limit values “Stg3” and “Stb3” of green (G). In the case of an image of sunrise, for example, the upper limit value “Stb3” of the first small area (small area Ry3) of blue (B) is the upper limit value of other colors, that is, red (R) or It is much larger than the upper limit values “Str3” and “Stg3” of green (G). On the other hand, in the middle tone, that is, here, the upper limit values “Str2”, “Stg2”, and “Stb2” of each color of the second small area (small area Ry2) do not have a property that is biased to a specific color.

  From this, the maximum difference ΔHt1 between the upper limit values “Str3”, “Stg3”, “Stb3” of each color of the first small area (small area Ry3) and each color of the second small area (small area Ry2) The maximum difference ΔHt2 between “St2”, “Stg2”, and “Stb2” is obtained, and the two differences ΔHt1 and ΔHt2 are compared to determine whether the image is sunset or sunrise. It can be easily determined.

  Here, since the relationship between the upper limit values “Str3”, “Stg3”, and “Stb3” of each color of the first small area (small area Ry3) is Str3> Stb3> Stg3, the first small area (small area Ry3) ), The maximum difference ΔHt1 between the upper limit values “Str3”, “Stg3” and “Stb3” of each color is the upper limit value “Str3” of red (R) and the upper limit value “Stg3” of green (G). It can be obtained from the difference. The maximum difference ΔHt2 between the upper limit values “Str2”, “Stg2”, and “Stb2” of each color in the second small area (small area Ry2) is the upper limit value of each color in the second small area (small area Ry2). Since the relationship of “Str2”, “Stg2”, and “Stb2” is Str2> Stb2> Stg2, the difference between the upper limit value “Str2” of red (R) and the upper limit value “Stg2” of green (G) Can be sought.

  On the other hand, if it is not an image of sunset or sunrise, as shown in FIG. 24B, highlights, that is, here, upper limit values “Str3”, “Stg3”, “Stg3”, “1st small region (small region Ry3)”. In “Stb3”, the upper limit value of a specific color does not become much larger than the upper limit values of other colors. Therefore, the maximum difference ΔHt1 between the upper limit values “Str3”, “Stg3”, and “Stb3” of each color in the first small area (small area Ry3) and the upper limit value of each color in the second small area (small area Ry2) The difference from the maximum difference ΔHt2 among “Str2”, “Stg2”, and “Stb2” does not become so large. From this, it can be easily determined that the image is not a sunset or sunrise image.

The maximum difference ΔHt1 between the colors “Str3”, “Stg3”, “Stb3” of each color in the first small region (small region Ry3), and the upper limit value “Str2” of each color in the second small region (small region Ry2) ”,“ Stg2 ”and“ Stb2 ”are compared with the maximum difference ΔHt2, for example, a difference between the difference ΔHt1 and the difference ΔHt2, that is, a value obtained by subtracting the difference ΔHt2 from the difference ΔHt1 is predetermined. It can be easily performed by checking whether or not the threshold value is equal to or greater than the threshold value “Kt”. That is, by checking whether or not the following relational expression (9) is satisfied, it can be easily determined whether the image is a sunset or sunrise image.
ΔHt1−ΔHt2 ≧ Kt (9)
By determining by the scanner driver by such a method, it is possible to accurately determine whether the image to be determined, that is, the image read by the image reading device 10 here is an image of sunset or sunrise. it can.

<Process flow>
FIG. 25 illustrates the flow of determination processing by the scanner driver. As shown in the figure, the scanner driver first sets the upper limit of each color of the first small area (small area Ry3) as attribute information regarding the first small area (small area Ry3) and the second small area (small area Ry2). The values “Str3”, “Stg3”, “Stb3” and the upper limit values “Str2”, “Stg2”, “Stb2” of each color of the second small area (small area Ry2) are acquired (S452). Next, based on the obtained upper limit values “Str2”, “Stg2”, “Stb2”, “Str3”, “Stg3”, “Stb3”, the scanner driver determines each color of the first small area (small area Ry3). Between the maximum difference ΔHt1 between the upper limit values “Str3”, “Stg3” and “Stb3” and the upper limit values “Str2”, “Stg2” and “Stb2” of each color in the second small area (small area Ry2) The maximum difference ΔHt2 is obtained (S454). Next, the scanner driver compares the obtained two differences ΔHt1 and ΔHt2 (S456). As a result of the comparison, it is checked whether or not the difference between these two differences ΔHt1 and ΔHt2, that is, ΔHt1−ΔHt2 is equal to or greater than a predetermined threshold value “Kt” (S458). If ΔHt1−ΔHt2 is equal to or greater than the predetermined threshold “Kt”, the scanner driver determines that the image is sunset or sunrise (S460). On the other hand, if ΔHt1−ΔHt2 is not equal to or greater than the predetermined threshold “Kt”, the scanner driver determines that the image is not a sunset or sunrise image (S462). After determining whether the image is a sunset or sunrise image in this way, the scanner driver immediately ends the process.

=== Normal Backlight Image Correction ===
Next, “normal correction processing for a backlight image” will be described. This correction process is performed when the image is determined not to be a backlight image in the “determination process for determining whether the image is a backlight image” and the image for sunset or sunrise is determined in the process for determining whether the image is a sunset or sunrise image. It is executed when it is determined that it is not.

  This correction processing is performed using the histograms for each of the RGB colors generated in the previous “backlight image determination”. That is, from each of the three small areas Ry1, Ry2, Ry3, Ry4 obtained by dividing in the histogram of each color of RGB generated in the previous “backlight image determination”, from the three small areas Ry1, Ry2, Ry3. Attribute information is acquired, and correction processing is performed on the image based on the acquired attribute information. The correction method will be described in detail below.

<Acquisition of attribute information>
First, the scanner driver performs three small areas Ry1, Ry2, Ry1, Ry3, Ry4 out of the four small areas Ry1, Ry2, Ry3, Ry4 obtained by segmenting in the histogram for each of the RGB colors generated in the above-mentioned “backlight image determination”. , Ry3 respectively to obtain attribute information. Here, as the attribute information, the average values of the density values of the pixels occupying the three small regions Ry1, Ry2, and Ry3 are acquired from the RGB histograms. Here, the average value of the density values of the pixels occupying the small area Ry1 of the red (R) histogram is “AVr1,” and the average value of the density values of the pixels occupying the small area Ry2 is “AVr2.” The average value of the density values of the pixels occupying the small area Ry3 is “AVr3”. In addition, the average value of the density values of the pixels occupying the small area Ry1 of the green (G) histogram is “AVg1,” and the average value of the density values of the pixels occupying the small area Ry2 is “AVg2.” The average value of the density values of the pixels occupying the region Ry3 is “AVg3”. Further, the average value of the density values of the pixels occupying the small area Ry1 of the blue (B) histogram is “AVb1,” and the average value of the density values of the pixels occupying the small area Ry2 is “AVb2.” The average value of the density values of the pixels occupying the region Ry3 is “AVb3”.

<Generation of correction information>
Thereafter, the scanner driver performs backlight correction processing on the image read by the image reading device 10 based on the average values AVr1, AVr2, AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3 acquired here. Generate correction information. Hereinafter, a correction information generation method according to the present embodiment will be described.

(1) Acquisition of information regarding luminance In this embodiment, each color of pixels occupying three small regions Ry1, Ry2, and Ry3 obtained from a histogram of three colors of red (R), green (G), and blue (B). Based on the average values AVr1, AVr2, AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3 of the different density values, the scanner driver assigns each of the small regions Ry1, Ry2, and Ry3 to the respective regions Ry1, Ry2, and Ry3. Average values Yav1, Yav2, and Yav3 of the luminance of the occupied pixels are acquired. Here, the scanner driver acquires the average values Yav1, Yav2, and Yav3 of the luminance of the pixels that occupy the small regions Ry1, Ry2, and Ry3 by using the following equations (10) to (12), for example.
Yav1 = 1/5 × AVr1 + 3/5 × AVg1 + 1/5 × AVb1 (10)
Yav2 = 1/5 × AVr 2 + 3/5 × AVg 2 + 1/5 × AVb 2 (11)
Yav3 = 1/5 × AVr3 + 3/5 × AVg3 + 1/5 × AVb3 (12)
Here, in the formulas (10) to (12), the case where the ratio of each color of red (R), green (G), and blue (B) is “1: 3: 1” is shown as an example. However, other ratios may be set.

In this way, the scanner driver obtains the average luminance values Yav1, Yav2, and Yav3 of the pixels that occupy the three small regions Ry1, Ry2, and Ry3. The scanner driver then calculates the average luminance of the entire image read by the image reading device 10 from the average luminance values Yav1, Yav2, and Yav3 of the pixels occupying the three small regions Ry1, Ry2, and Ry3. The value Yav0 is obtained. In this embodiment, since the three small areas Ry1, Ry2, and Ry3 occupy 99% of the entire image, the average luminance values Yav1, Yav2, and Yav3 corresponding to the three small areas Ry1, Ry2, and Ry3, respectively. Is an average value Yav0 of the overall luminance of the image. Therefore, in the present embodiment, the average value Yav0 of the overall luminance of the image read by the image reading device 10 can be obtained by the following equation (13), for example.
Yav0 = (Yav1 + Yav2 + Yav3) / 3 (3)
In this way, the scanner driver obtains the average value Yav0 of the overall luminance of the image read by the image reading device 10.

(2) Acquisition of Target Value The scanner driver acquires the average values Yav1, Yav2, and Yav3 of the pixels that occupy the three small regions Ry1, Ry2, and Ry3 in this way, and further acquires the image read by the image reading device 10 After obtaining the average value Yav0 of the overall luminance, the target value is obtained. This target value is a value acquired for performing the backlight correction process on the image read by the image reading device 10, and is a value that serves as an index in determining the outline of the backlight correction process. In the present embodiment, two target values Ym1 and Ym2 are acquired as the target values. One target value Ym1 is acquired corresponding to the small region Ry1, and indicates the target value of the average value of the luminance of the pixels occupying the small region Ry1. The other target value Ym2 is acquired corresponding to the small area Ry2, and indicates the target value of the average value of the luminance of the pixels occupying the small area Ry2.

  Next, a method for obtaining these two target values Ym1 and Ym2 will be described. FIG. 26 illustrates an example of a method for obtaining two target values Ym1 and Ym2.

  In the present embodiment, the scanner driver obtains these two target values Ym1 and Ym2 based on the average luminance value Yav1 of the pixels occupying the small region Ry1 and the average luminance value Yav0 of the entire image. Specifically, the scanner driver adds the average value Yav1 of the luminance of the pixels occupying the small area Ry1 and the average luminance Yav0 of the entire luminance of the image, and based on the addition value Sadd obtained thereby, The values Ym1 and Ym2 are acquired. In obtaining the two target values Ym1 and Ym2 from the added value Sadd, two relational expressions (A) and (B) as shown in the figure are used.

  The two relational expressions (A) and (B) shown in the figure are represented on a graph formed with the horizontal axis as an addition value and the vertical axis as a target value. The relational expression (A) is for obtaining the target value Ym1 corresponding to the small region Ry1. The relational expression (B) is for obtaining the target value Ym2 corresponding to the small region Ry2.

  Assume that the scanner driver adds the average luminance value Yav1 of the pixels occupying the small area Ry1 and the average luminance value Yav0 of the entire luminance of the image to obtain the added value Sadd. Then, the scanner driver acquires the target value Ym1 corresponding to the addition value Sadd from the relational expression (A) based on the addition value Sadd. Also, the scanner driver obtains the target value Ym2 corresponding to the addition value Sadd from the relational expression (B) based on the addition value Sadd. As described above, the scanner driver can obtain two target values, that is, the addition value Sadd obtained by adding the average value Yav1 of the luminance of the pixels occupying the small region Ry1 and the average value Yav0 of the entire luminance of the image. It is possible to easily obtain the target value Ym1 of the average value of the luminance of the pixels occupying the small region Ry1 and the target value Ym2 of the average value of the luminance of the pixels occupying the small region Ry2.

  Here, the reason why the scanner driver has acquired the two target values Ym1 and Ym2 based on the addition value Sadd will be described. The small region Ry1 from which the average value Yav1 is acquired is the small region having the smallest density value among the four small regions obtained by dividing the histogram generated for each color. For this reason, the magnitude of the average value Yav1 acquired from the small region Ry1 varies greatly depending on the degree of backlight of the image. However, it is impossible to determine whether or not the image is a backlight image only by the average value Yav1 acquired from the small region Ry1. This is because even if the entire image is a dark image, the average value Yav1 acquired from the small region Ry1 is small. In this case, it is necessary to increase the brightness of the entire image. This is because a sufficient correction effect cannot be obtained only by increasing the luminance of some pixels as in a backlight image.

  Therefore, in order to consider the average value Yav0 of the entire luminance of the image, the average value Yav0 of the entire luminance of the image is added to the average value Yav1 of the pixels occupying the small area Ry1, so that the addition value Sadd is obtained. is there. This added value Sadd is a very appropriate value in examining the degree of backlighting of the image. That is, when the added value Sadd is very small, not only the average luminance value Yav1 of the pixels occupying the small region Ry1 but also the average luminance value Yav0 of the entire image is small and the entire image is dark. It can be judged that there is. As a result, correction can be performed to increase the brightness of the entire image.

  On the other hand, when the added value Sadd is not so small, the average luminance value Yav0 of the entire image is not so small, and the average value Yav1 of the pixels occupying the small area Ry1 is small. Judgment can be made. Accordingly, it is possible to perform correction so as to increase the luminance of some pixels, not the luminance of the entire image.

  The two relational expressions (A) and (B) described above are set so that the two target values Ym1 and Ym2 are defined according to the magnitude of the added value Sadd. That is, the two relational expressions (A) and (B) increase the luminance of the entire image when the added value Sadd is very small, that is, when the luminance of the entire image is low and it is determined that the image is a dark image. In order to correct in this way, the target value Ym1 of the average value of the luminance of the pixels occupying the small region Ry1 and the target value Ym2 of the average value of the luminance of the pixels occupying the small region Ry2 are respectively very large values. Is set. When the addition value Sadd is not so small, that is, when it is determined that the image is a backlight image, the two relational expressions (A) and (B) are the average values of the luminances of the pixels occupying the small region Ry1. The target value Ym1 and the target value Ym2 of the average luminance of the pixels occupying the small area Ry2 are set so as not to be too large. Accordingly, the “correction processing for a normal backlight image” of the present embodiment is appropriate even when the image read by the image reading device 10 is not a backlight image but is a dark image with low overall brightness. Correction can be performed.

  These two relational expressions (A) and (B) can be set as follows, for example. That is, assuming that the minimum value that the addition value Sadd can take is “Smin” and the maximum value that the addition value Sadd can take is “Smax”, the maximum target value corresponding to each of the minimum value “Smin” and the maximum value “Smax”. Set the minimum target value. Here, the maximum target value corresponding to the relational expression (A) is “Ymax1”, and the minimum target value is “Ymin1”. The maximum target value corresponding to the relational expression (B) is “Ymax2”, and the minimum target value is “Ymin2”. The minimum value “Smin” and the maximum value “Smax” of the added value Sadd, the maximum target value “Ymax1” and the minimum target value “Ymin1” of the relational expression (A), and the maximum target value “of the relational expression (B)” Ymax2 ”and the minimum target value“ Ymin2 ”are appropriately set based on empirical rules and the like.

  From this, the relational expression (A) is, for example, the point Pm1 defined by the minimum value “Smin” of the addition value Sadd and the maximum target value “Ymax1” of the relational expression (A) as shown in FIG. And a straight line connecting the maximum value “Smax” of the addition value Sadd and the point Pm3 defined by the minimum target value “Ymin1” of the relational expression (A). Further, the relational expression (B) is, for example, as shown in the figure, a point Pm2 defined by the minimum value “Smin” of the addition value Sadd and the maximum target value “Ymax2” of the relational expression (B), It can be easily set as a straight line connecting the maximum value “Smax” of the addition value Sadd and the point Pm4 defined by the minimum target value “Ymin2” of the relational expression (B).

  Here, the minimum target value “Ymin1” of the relational expression (A) corresponding to the maximum value “Smax” of the addition value Sadd and the minimum target value “Ymin2” of the relational expression (B) are respectively set to very small values. Yes. Therefore, when the added value Sadd is too large, the target value Ym1 of the average value of the luminance of the pixels occupying the small region Ry1 and the target value Ym2 of the average value of the luminance of the pixels occupying the small region Ry2 are standard. It can be set to a value sufficiently smaller than the correct value. As a result, even when the brightness of the entire image is high, that is, when the entire image is bright, the image can be corrected. That is, in the present embodiment, an appropriate backlight correction process can be performed on an image without determining whether the image is a backlight image, that is, regardless of whether the image is a backlight image. Therefore, in the present embodiment, even if it is determined that the image is not a backlight image in the “determination process for determining whether the image is a backlight image”, the “correction process for a normal backlight image” can be performed. .

  In this way, the scanner driver calculates the small region Ry1 from the relational expressions (A) and (B) based on the average luminance value Yav1 of the pixels occupying the small region Ry1 and the average luminance value Yav0 of the entire image. The target value Ym1 of the average value of the luminance of the occupied pixels and the target value Ym2 of the average value of the luminance of the pixels occupying the small region Ry2 can be easily obtained.

  In the present embodiment, based on the added value Sadd obtained by adding the average luminance value Yav1 of the pixels occupying the small area Ry1 and the average luminance value Yav0 of the entire image, two target values Ym1, In order to easily obtain Ym2, a straight line, that is, a linear function is used in the relational expressions (A) and (B), but in the relational expressions (A) and (B) used here, However, such a linear function is not necessarily used. That is, a nonlinear function may be used in relational expressions (A) and (B).

  In the present embodiment, the average value Yav1 of the luminance of the pixels occupying the small region Ry1 and the average luminance Yav0 of the entire image are added, and the target value Ym1 is obtained based on the addition value Sadd obtained thereby. , Ym2 has been acquired. However, in order to acquire the target values Ym1, Ym2, it is not always necessary to obtain such an added value Sadd. That is, the target value may be acquired based only on the average luminance Yav1 of the pixels occupying the small region Ry1, or the target value may be acquired based on the average luminance Yav2 of the pixels occupying the small region Ry2. Also good.

  In this embodiment, the two target values Ym1 and Ym2 are acquired based on the addition value Sadd. However, it is not always necessary to acquire the two target values Ym1 and Ym2, and the number of target values is as follows. There may be one, or three or more.

(3) Generation of correction information The scanner driver generates correction information based on the two target values Ym1 and Ym2 acquired in this way. This correction information is information for the scanner driver to perform backlight correction processing on the image read by the image reading device 10. In the present embodiment, the scanner driver executes the above-described “density correction” as the backlight correction process. Therefore, in this embodiment, the scanner driver generates setting information for “density correction” as correction information. A method for generating setting information in “density correction” will be described in detail below.

FIG. 27 illustrates a method for generating setting information in “density correction” as correction information. Here, the scanner driver performs a backlight correction process on the image read by the image reading device 10 by adjusting the tone curve Tc of “density correction”.
In this embodiment, the tone curve Tc corresponding to all the RGB (red, green, and blue) colors described with reference to FIG. 8 is adjusted as the tone curve Tc for “density correction”.

  Therefore, the scanner driver sets arbitrary three points Ps1, Ps2, and Ps3 through which the tone curve Tc passes, based on the two target values Ym1 and Ym2, as correction information, as shown in FIG. Then, the scanner driver adjusts the tone curve Tc so as to pass through these three points Ps1, Ps2, and Ps3. In the present embodiment, the point Ps1 is set corresponding to the small region Ry1. The point Ps2 is set corresponding to the small area Ry2. The point Ps3 is set corresponding to the small area Ry3. In this embodiment, the input value of the point Ps1 is “X1”, and the output value is “Y1”. The input value of the point Ps2 is “X2”, and the output value is “Y2”. The input value of the point Ps3 is “X3”, and the output value is “Y3”.

  Here, the input values “X1”, “X2”, and “X3” of the three points Ps1, Ps2, and Ps3 are values that the input value can take, that is, in the present embodiment, a range of “0” to “255”. Any suitable value may be set. However, in order to form a well-balanced tone curve Tc, it is preferable to set as follows. That is, since these three points Ps1, Ps2, and Ps3 are arranged equally in a balanced manner, the range of values that the input value can take, that is, “0” to “255” is divided into three, The input values “X1”, “X2”, and “X3” are arranged respectively. That is, the input value “X1” is set within the range of “0” to “84”. The input value “X2” is set within the range of “85” to “169”. Further, the input value “X3” is set within the range of “170” to “255”.

  Furthermore, for example, it is preferable to set the input values “X1”, “X2”, and “X3” so as to be positioned at the approximate center in each section. Specifically, in this embodiment, the input value “X1” of the point Ps1 is set to “42”. Further, the input value “X2” of the point Ps2 is set to “127”. Further, the input value “X3” of the point Ps3 is set to “212”.

  On the other hand, the output values “Y1”, “Y2”, and “Y3” of these three points Ps1, Ps2, and Ps3 are set as follows in this embodiment. That is, the output value “Y1” of the point Ps1 and the output value “Y2” of the point Ps2 are set based on the two target values Ym1 and Ym2. On the other hand, the output value “Y3” of the point Ps3 is set based on the point Ps2 and the point Ps0. Here, the point Ps0 is a point set with an input value “255” and an output value “255”. The point Ps3 is set so as to be positioned almost directly above the straight line set between the point Ps2 and the point Ps0. That is, when “X3” is obtained as the input value of the point Ps3, a point corresponding to the input value “X3” is searched for on the straight line connecting the point Ps2 and the point Ps0 and set as the point Ps3. it can. The output value “Y3” of the point Ps3 can be obtained from the straight line connecting the point Ps2 and the point Ps0 as the output value corresponding to the input value “X3”.

  Here, how to obtain the output values “Y1” and “Y2” of the points Ps1 and Ps2 will be described. First, how to obtain the output value “Y1” of the point Ps1 will be described. The point Ps1 is set corresponding to the small area Ry1. The output value “Y1” of the point Ps1 is set so that the average value Yav1 of the luminance of the pixels occupying the small area Ry1 of the corrected histogram becomes the target value Ym1. Specifically, this is performed as follows. That is, an output value obtained when “density correction” is not performed on the input value “X1” is “Yt1”. That is, the output value “Yt1” is the same value as the input value “X1”. In FIG. 27, the point when “density correction” is not performed on the input value “X1” is indicated as “Pt1”. A difference ΔYp1 between the output value “Y1” of the point Ps1 and the output value “Yt1” is obtained. This difference ΔYp1 can be obtained as follows.

  FIG. 28A illustrates a method for obtaining the difference ΔYp1 according to the present embodiment. In the present embodiment, the difference ΔYp1 is obtained by the scanner driver. First, as shown in the figure, the scanner driver obtains an output value “Yt1” corresponding to the input value “X1” (S502). The output value “Yt1” is the same value as the input value “X1” here. Next, the scanner driver sets the difference ΔYp1 to an initial value, here “0” (S504). Then, the scanner driver sets a value obtained by adding “1” to the difference ΔYp1 as a new ΔYp1 (S506). That is, when ΔYp1 is “0”, the new ΔYp1 is added by “1” to become “1”.

  Next, the scanner driver sets a value obtained by adding ΔYp1 to the previously acquired output value “Yt1” as the output value “Y1” of the point Ps1 (S508). Then, image data after correction when “density correction” is executed based on the tone curve Tc set by the point Ps1 is acquired (S510). Here, as described with reference to FIG. 11, the contents of various adjustments (corrections) such as “histogram adjustment” and “image adjustment” are also reflected in addition to “density correction”. And

  After performing the correction in this manner, the scanner driver generates a histogram based on the image data obtained by the correction (S512). Here, the scanner driver creates a histogram of three colors of red (R), green (G), and blue (B). Next, the scanner driver divides the histogram created in this way into small regions (S514). In other words, as described above, the scanner driver divides the region represented by the histogram into four small regions Ry1, Ry2, Ry3, and Ry4.

  Then, the scanner driver acquires attribute information related to the small region Ry1 from the four regions Ry1, Ry2, Ry3, and Ry4 obtained in this way. Here, the scanner driver obtains average values AVr1, AVg1, and AVb1 of the density values of each color of red (R), green (G), and blue (B) as attribute information related to the small region Ry1. Then, the scanner driver obtains the average value Yav1 of the luminance of the pixels occupying the small region Ry1 based on the acquired average values AVr1, AVg1, and AVb1 of the density values of the respective colors (S516). Here, the scanner driver obtains the average value Yav1 by, for example, Expression (7).

  Thereafter, the scanner driver compares the obtained average value Yav1 with the previously acquired target value Ym1 (S518). Here, the scanner driver checks whether or not the obtained average value Yav1 is equal to or greater than the target value Ym1. If the calculated average value Yav1 is not equal to or greater than the target value Ym1, the scanner driver returns to step S506 and again adds “1” to the difference ΔYp1 to obtain a new ΔYp1. (S506). Then, the output value “Y1” of the point Ps1 is acquired again (S508), the image is corrected based on this (S510), a histogram is generated (S512), and the image is divided into small regions (S514). ), The average value Yav1 of the luminance of the pixels occupying the small region Ry1 is obtained (S516). Thereafter, the obtained average value Yav1 is compared with the target value Ym1, and it is checked again whether or not the obtained average value Yav1 has reached the target value Ym1 (S518). The scanner driver repeats such processing (S506 to S518) until the obtained average value Yav1 reaches the target value Ym1. That is, the difference ΔYp1 is incremented by “1” sequentially until the calculated average value Yav1 reaches the target value Ym1.

  When the obtained average value Yav1 reaches the target value Ym1, the scanner driver sets the output value “Y1” of the point Ps1 based on the difference ΔYp1 obtained here (S520). In other words, the scanner driver sets the value obtained by adding the difference ΔYp1 to the output value “Yt1” as the output value “Y1” of the point Ps1.

  In this way, the scanner driver sets the output value “Y1” of the point Ps1 such that the average value Yav1 of the luminance of the pixels occupying the small area Ry1 of the corrected histogram becomes the target value Ym1. When the output value “Y1” of the point Ps1 is set to a value smaller than the output value “Yt1”, the scanner driver determines in step S518 whether or not the average value Yav1 is lower than the target value Ym1. It is good to check.

  On the other hand, the output value “Y2” of the point Ps2 is set similarly to the output value “Y1” of the point Ps1. That is, the point Ps2 is set corresponding to the small region Ry2. Similarly to the case of the point Ps1, the output value “Y2” of the point Ps2 is set so that the average value Yav2 of the luminance of the pixels occupying the small area Ry2 of the corrected histogram becomes the target value Ym2. In short, if the output value obtained when “density correction” is not applied to the input value “X2” is “Yt2”, the difference ΔYp2 of the output value “Y2” of the point Ps2 with respect to this output value “Yt2” is obtained. Ask. In FIG. 27, a point when “density correction” is not performed on the input value “X2” is indicated as “Pt2”.

  FIG. 28B illustrates a method for obtaining the difference ΔYp2 according to the present embodiment. The difference ΔYp2 is also obtained by the scanner driver as in the case of the point Ps1. The scanner driver first obtains an output value “Yt2” corresponding to the input value “X2” as shown in FIG. The output value “Yt2” is the same value as the input value “X2” here. Next, the scanner driver sets the difference ΔYp2 to an initial value, here “0” (S554). Then, the scanner driver sets a value obtained by adding “1” to the difference ΔYp2 as a new ΔYp2 (S556).

  Next, the scanner driver sets a value obtained by adding ΔYp2 to the previously acquired output value “Yt2” as the output value “Y2” of the point Ps2 (S558). Then, image data after correction when “density correction” is executed based on the tone curve Tc set by the point Ps2 is acquired (S560). Here, as described with reference to FIG. 11, the contents of various adjustments (corrections) such as “histogram adjustment” and “image adjustment” are also reflected in addition to “density correction”. And Here, the scanner driver may or may not reflect the setting content of the point Ps1.

  After correcting in this way, the scanner driver generates a histogram based on the image data obtained by this correction (S562). Here, the scanner driver creates a histogram of three colors of red (R), green (G), and blue (B). Next, the scanner driver divides the histogram created in this way into four small regions Ry1, Ry2, Ry3, and Ry4 (S564).

  Then, the scanner driver acquires attribute information related to the small region Ry2 from the four regions Ry1, Ry2, Ry3, and Ry4 obtained in this way. Here, the scanner driver obtains the average values AVr2, AVg2, and AVb2 of the density values of each color of red (R), green (G), and blue (B) as attribute information related to the small region Ry2. Then, the scanner driver obtains the average value Yav2 of the luminance of the pixels occupying the small area Ry2 based on the acquired average values AVr2, AVg2, and AVb2 of the density values of the respective colors (S566). Here, the scanner driver obtains the average value Yav2 by, for example, Equation (8).

  Thereafter, the scanner driver compares the obtained average value Yav2 with the previously obtained target value Ym2 (S568). Here, if the obtained average value Yav2 has not reached the target value Ym2, the scanner driver returns to step S556, adds “1” again to obtain a new ΔYp2 (S556), and The output value “Y2” of the point Ps2 is acquired (S558), and based on this, the image is corrected (S560). Then, the scanner driver generates a histogram of the corrected image (S562), divides it into four small regions (S564), and obtains the average value Yav2 again (S566). Thereafter, the acquired average value Yav2 is compared with the target value Ym2, and it is checked again whether or not the obtained average value Yav2 has reached the target value Ym2 (S568). The scanner driver repeats such processing (S556 to S568) until the obtained average value Yav2 reaches the target value Ym2. That is, the difference ΔYp2 is incremented by “1” sequentially until the obtained average value Yav2 reaches the target value Ym2.

  When the obtained average value Yav2 reaches the target value Ym2, the scanner driver sets the output value “Y2” of the point Ps2 based on the difference ΔYp2 obtained here (S570). In other words, the scanner driver sets the value obtained by adding the difference ΔYp2 to the output value “Yt2” as the output value “Y2” of the point Ps2.

  In this way, the scanner driver sets the output value “Y2” of the point Ps2 such that the average value Yav2 of the luminance of the pixels occupying the small area Ry2 of the corrected histogram becomes the target value Ym2. When the output value “Y2” of the point Ps2 is set to a value smaller than the output value “Yt2”, the scanner driver determines in step S568 whether or not the average value Yav2 is equal to or less than the target value Ym2. It is good to check.

  Note that the output value “Y3” of the point Ps3 can be obtained from the straight line connecting the point Ps2 and the point Ps0 as the output value corresponding to the input value “X3” as described above.

(4) Other setting methods of the input values X1, X2, and X3 Other setting methods of the input values “X1”, “X2”, and “X3” of the three points Ps1, Ps2, and Ps3 described above are as follows: There are methods. That is, for example, the average luminance values Yav1, Yav2, and Yav3 of the pixels that occupy the small areas Ry1, Ry2, and Ry3 may be set. That is, the input values “X1”, “X2”, and “X3” of these three points Ps1, Ps2, and Ps3 are reflected in the average luminance values Yav1, Yav2, and Yav3 of the pixels that occupy the small regions Ry1, Ry2, and Ry3. Specifically, there are the following setting methods.

  29A to 29C reflect three input values “X1”, “X2”, and “X3”, reflecting the average luminance values Yav1, Yav2, and Yav3 of the pixels that occupy the three small regions Ry1, Ry2, and Ry3. An example of setting is described. FIG. 29A illustrates a method for setting the input value “X1” of the point Ps1. FIG. 29B illustrates a method for setting the input value “X2” of the point Ps2. FIG. 29C illustrates a method of setting the input value “X3” of the point Ps3.

  When the input value “X1” of the point Ps1 is set, as shown in FIG. 29A, a predetermined target range “a1” within the range of “0” to “84” in which the input value “X1” is set. ~ “B1” is set. The predetermined target ranges “a1” to “b1” are set to approximately the center between “0” to “84”. “A1” and “b1” are set to appropriate values between “0” and “84”, respectively. The predetermined target ranges “a1” to “b1” set in this way are divided into 256 equal parts, and values “0” to “255” are allocated, and the average value Yav1 of the luminance of the pixels occupying the small region Ry1 is obtained. The corresponding point Pk1 is specified. Then, a value between “0” and “84” corresponding to the point Pk1 is obtained. The obtained value is set as the input value “X1” of the point Ps1.

  When the input value “X2” of the point Ps2 is set, as shown in FIG. 29B, the predetermined target range “8” is set within the range of “85” to “169” where the input value “X2” is set. “a2” to “b2” are set. The predetermined target ranges “a2” to “b2” are set to approximately the center between “85” to “169”. “A2” and “b2” are set to appropriate values between “85” and “169”, respectively. The predetermined target ranges “a2” to “b2” set in this way are divided into 256 equal parts, and values “0” to “255” are allocated, and the average value Yav2 of the luminance of the pixels occupying the small area Ry2 is obtained. The corresponding point Pk2 is specified. Then, a value between “85” and “169” corresponding to the point Pk2 is obtained. The obtained value is set as the input value “X2” of the point Ps2.

  When the input value “X3” of the point Ps3 is set, as shown in FIG. 29C, a predetermined target range “in the range of“ 170 ”to“ 255 ”where the input value“ X3 ”is set. “a3” to “b3” are set. The predetermined target ranges “a3” to “b3” are set to approximately the center between “170” to “255”. “A3” and “b3” are set to appropriate values between “170” and “255”, respectively. The predetermined target ranges “a3” to “b3” set in this way are divided into 256 equal parts, and values “0” to “255” are allocated, and the average value Yav3 of the luminance of the pixels occupying the small region Ry3 is obtained. The corresponding point Pk3 is specified. Then, a value between “170” and “255” corresponding to the point Pk3 is obtained. The obtained value is set as the input value “X3” of the point Ps3.

  In this way, the input values “X1”, “X2”, “X3” of the three points Ps1, Ps2, and Ps3 are reflected reflecting the average values Yav1, Yav2, and Yav3 of the luminance of the pixels that occupy the small regions Ry1, Ry2, and Ry3. ”Is set for predetermined reasons“ a1 ”to“ b1 ”,“ a2 ”to“ b2 ”, and“ a3 ”to“ b3 ”for the following reason. That is, when the average values Yav1, Yav2, and Yav3 of the pixels occupying the small regions Ry1, Ry2, and Ry3 are set as input values “X1”, “X2”, and “X3” as they are, the input value “X1” This is because the range of values that can be taken by “X2” and “X3” is widened, and thus there is a possibility that the tone curve Tc to be adjusted may be greatly affected.

  In other words, for example, it is assumed that the average luminance value Yav1 of the pixels occupying the small region Ry1 is very small. When the average value Yav1 is set as the input value “X1” of the point Ps1 as it is and the point Ps1 is set in the tone curve Tc on FIG. 27, the position of the point Ps1 moves to the left side in the figure and moves to the vertical axis. Because of the approach, the rise of the tone curve Tc becomes very steep. When the average luminance value Yav1 of the pixels occupying the small region Ry1 is very large, the position of the point Ps1 moves to the right side in the drawing and approaches the point Ps2, so that the shape of the tone curve Tc is distorted. There is a risk of it. The same applies to points Ps2 and Ps3.

  Therefore, in order to prevent the positions of the points Ps1, Ps2, and Ps3 from being too close to each other, each value that defines a predetermined target range, that is, “a1”, “b1”, “a2”, “b2”, By setting “a3” and “b3” to appropriate values, the shape of the tone curve Tc can be prevented from being distorted.

  In this way, the scanner driver sets three points Ps1, Ps2, and Ps3 for adjusting the tone curve Tc in “density correction” as correction information based on the two target values Ym1 and Ym2 acquired. Generate information.

  In the present embodiment, the three points Ps1, Ps2, and Ps3 are set as the correction information. However, the three points Ps1, Ps2, and Ps3 are not necessarily set as the correction information. That is, the number of points set as correction information may be one or two, and may be four or more. The number of points to be set is appropriately set according to the number of small regions obtained by dividing the region represented by the three-color histogram of red (R), green (G), and blue (B). Alternatively, it may be set as appropriate according to the number of small areas selected to acquire the attribute information.

  In addition, for the three points Ps1, Ps2, and Ps3 set as correction information, ranges of values that can be taken by the input value, that is, “0” to “255” in this embodiment are divided into three. It does not have to be arranged in each section.

  In the present embodiment, setting information of three points Ps1, Ps2, and Ps3 for adjusting the tone curve Tc in “density correction” is generated as correction information because of the relationship in which “density correction” is executed as the backlight correction processing. However, it is not always necessary to generate such setting information as correction information. That is, appropriate information may be generated as correction information in accordance with processing executed as backlight correction processing.

<Backlight correction processing>
The scanner driver is read by the image reading device 10 based on the information regarding the three points Ps1, Ps2, and Ps3 for adjusting the tone curve Tc in the “density correction” generated as correction information in this way. Backlight correction processing is performed on the image. Specifically, in the “density correction” in the adjustment procedure described with reference to FIG. 11, the image reading device is based on the tone curve Tc adjusted based on the information about the three points Ps1, Ps2, and Ps3 generated as the correction information. The density value of each color of the data of each pixel constituting the image read by 10 is converted. As a result, the scanner driver performs backlight correction processing on the image read by the image reading device 10.

  Here, the image subjected to the backlight correction processing is an image that has already been subjected to histogram adjustment, image adjustment (excluding (3) saturation adjustment), and the like according to the adjustment procedure shown in FIG. The scanner driver performs backlight correction processing by “density correction” on an image that has already been subjected to histogram adjustment, image adjustment (excluding (3) saturation adjustment), and the like in this way.

=== Correction of Backlight Image with Halftone Color Balance Loss ===
Next, “correction processing for a backlight image in which the halftone color balance is lost” will be described. This correction process is performed when it is determined in the “determination process whether or not the image has a halftone color balance collapsed” that the image has a halftone color balance collapsed.

  This correction process is performed according to the procedure described with reference to FIG. However, here, the scanner driver performs correction processing using the histogram of each color of RGB generated in the above-mentioned “backlight image determination”. That is, as in the case of “correction processing for a normal backlight image”, four small regions Ry1, Ry2, Ry3 obtained by segmentation in the histogram of each color of RGB generated in the previous “determination of backlight image”, Attribute information is acquired from each of the three small areas Ry1, Ry2, and Ry3 of Ry4, and “correction processing for a backlight image in which a halftone color balance is lost” is performed based on the acquired attribute information. Correction information is generated, and an image is corrected based on the correction information. The correction method will be described in detail below.

<Acquisition of attribute information>
First, the scanner driver performs three small areas Ry1, Ry2, Ry1, Ry3, Ry4 out of the four small areas Ry1, Ry2, Ry3, Ry4 obtained by segmenting in the histogram for each of the RGB colors generated in the above-mentioned “backlight image determination”. , Ry3 respectively to obtain attribute information. Here, as the attribute information, the average values AVr1, AVr2, and AVr3 of the density values of the pixels that occupy the three small regions Ry1, Ry2, and Ry3 from the histogram of each color of RGB are the same as in the case of “normal backlight image correction”. , AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3 are obtained.

<Generation of correction information>
After that, the scanner driver performs correction for correcting the backlight image in which the halftone color balance is lost based on the acquired average values AVr1, AVr2, AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3. Generate information. Hereinafter, a correction information generation method according to the present embodiment will be described.

(1) Acquisition of Information on Luminance Here, the scanner driver obtains the average values AVr1, AVr2, AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, AVb3 as in the case of “normal backlight image correction”. Based on the above, for the three small regions Ry1, Ry2, and Ry3, average values Yav1, Yav2, and Yav3 of the luminance of the pixels that occupy the regions Ry1, Ry2, and Ry3 are obtained. At this time, the scanner driver uses, for example, the above formulas (10) to (12). Further, the scanner driver obtains the average luminance value Yav0 of the entire image read by the image reading device 10 based on the acquired average luminance values Yav1, Yav2, and Yav3. At this time, the scanner driver uses the above equation (13), for example.

(2) Acquisition of Target Value Next, the scanner driver obtains the average luminance values Yav1, Yav2, and Yav3 of the pixels that occupy the three small regions Ry1, Ry2, and Ry3 and the average luminance value Yav0 of the entire image. Based on the above, the target value is obtained. This target value is a value acquired for executing the correction process on the backlight image in which the halftone color balance is lost, and is a value serving as an index in determining the outline of the correction process. In the present embodiment, two target values Ym1 and Ym2 are acquired as the target values in order to correct a backlight image whose halftone color balance has been lost. Among these, the target value Ym1 is acquired corresponding to the small region Ry1. Further, the target value Ym2 is acquired corresponding to the small region Ry2. Here, the small area Ry2 corresponds to a “selected small area”.

  These two target values Ym1 and Ym2 are obtained by adding the average value Yav1 of the luminance of the pixels occupying the small region Ry1 and the average value Yav0 of the entire luminance of the image as in the case of “normal backlight image correction”. Based on the added value Sadd obtained in this way, the relational expressions (A) and (B) shown in FIG. That is, the target value Ym1 is acquired as a value corresponding to the added value Sadd from the relational expression (A). The target value Ym2 is acquired as a value corresponding to the added value Sadd from the relational expression (B).

(3) Generation of correction information The scanner driver generates correction information based on the two target values Ym1 and Ym2 acquired in this way. This correction information is information for performing an appropriate correction process on the backlight image in which the halftone color balance is lost. In this embodiment, the scanner driver performs the above-described “density correction” as the correction process for the backlight image in which the halftone color balance is lost, as in the case of “normal backlight image correction”. Therefore, in this embodiment, the scanner driver generates setting information for “density correction” as correction information. A method for generating the setting information will be described in detail below.

  The scanner driver performs each “red (R)”, “green (G)”, and “blue (B)” color described with reference to FIG. 8 in order to execute “density correction” on the backlight image whose halftone color balance is lost. Tone curves Tcr, Tcg, and Tcb are individually adjusted. Here, the scanner driver adjusts the tone curves Tcr, Tcg, and Tcb of each color of red (R), green (G), and blue (B) based on the acquired two target values Ym1 and Ym2. The two target values Ym1 and Ym2 are used as target values common to the respective colors when adjusting the tone curves Tcr, Tcg, and Tcb of the respective RGB colors. As a method for adjusting the tone curves Tcr, Tcg, and Tcb of each of the RGB colors based on the two target values Ym1 and Ym2, the same method as in the case of “normal backlight image correction” described in FIG. 27 is adopted. .

  FIG. 30 illustrates a method of adjusting the red (R) tone curve Tcr based on the two target values Ym1 and Ym2. Here, a method for adjusting the red (R) tone curve Tcr will be described as an example.

  As shown in the figure, the scanner driver sets three points Pr1, Pr2, and Pr3 through which the tone curve Tcr passes based on the two target values Ym1 and Ym2 as correction information. The point Pr1 is set corresponding to the small area Ry1, the point Pr2 is set corresponding to the small area Ry2, and the point Pr3 is set corresponding to the small area Ry3. The input values of the points Pr1, Pr2, and Pr3 are “Xr1”, “Xr2”, and “Xr3”, respectively. In addition, the output values of the points Pr1, Pr2, and Pr3 are “Yr1”, “Yr2”, and “Yr3”, respectively. Here, the input values “Xr1”, “Xr2”, and “Xr3” of the points Pr1, Pr2, and Pr3 are the input values “X1”, “X2”, and “X3” of the points Ps1, Ps2, and Ps3 described in FIG. Set in the same way as

  On the other hand, the output values “Yr1” and “Yr2” of the points Pr1 and Pr2 are set based on the two target values Ym1 and Ym2. The output value “Yr3” of the point Pr3 is set on a straight line connecting the point Pr2 and the point Pr0. The point Pr0 is a point where the input value is “255” and the output value is “255”. If the input value “Xr3” of the point Pr0 is determined, the output value “Yr3” can be obtained.

  The output values “Yr1” and “Yr2” of the points Pr1 and Pr2 are obtained by the same method as in the case of “normal backlight image correction”. That is, for example, in the case of the output value “Yr1” of the point Pr1, the point value is set so that the average value AVr1 of the density values of the pixels occupying the small region Ry1 of the corrected red (R) histogram becomes the target value Ym1. The output value “Yr1” of Pr1 is set. Specifically, the output value “Yrt1” (the same value as the input value “Xr1”) obtained when the “density correction” is not performed on the input value “Xr1” is a point for the output value “Yrt1”. The difference ΔYrp1 of the output value “Yr1” of Pr1 is obtained.

  FIG. 31A explains a method for obtaining this difference ΔYrp1. The scanner driver first obtains the output value “Yrt1” from the input value “Xr1” as shown in FIG. Next, the difference ΔYrp1 is set to an initial value “0” (S604). Then, a value obtained by adding “1” to the difference ΔYrp1 is set as a new ΔYrp1 (S606). Next, a value obtained by adding ΔYrp1 to the previously obtained output value “Yrt1” is set as the output value “Yr1” of the point Pr1 (S608), and “density correction” is performed based on the tone curve Tcr set by this point Pr1. "Is acquired (S610). Here, as described with reference to FIG. 11, the contents of various adjustments (corrections) such as “histogram adjustment” and “image adjustment” are also reflected in addition to “density correction”. And

  After correcting in this way, a histogram of red (R) is generated based on the image data obtained by this correction (S612). Then, the created red (R) histogram is divided into small regions (S614). Here, the histogram is divided into four small regions Ry1, Ry2, Ry3, and Ry4. Then, an average value AVr1 of density values is acquired as attribute information related to the small region Ry1 (S616). Thereafter, the obtained average value AVr1 is compared with the previously obtained target value Ym1 (S618), and it is checked whether or not the obtained average value AVr1 is equal to or greater than the target value Ym1. If the calculated average value AVr1 is not equal to or greater than the target value Ym1, the process returns to step S606, and again adds “1” to the difference ΔYrp1 to obtain a new ΔYrp1 (S606). The scanner driver repeatedly performs such processing (S606 to S618) until the obtained average value AVr1 reaches the target value Ym1. That is, the difference ΔYrp1 is incremented by “1” sequentially until the obtained average value AVr1 reaches the target value Ym1.

  When the calculated average value AVr1 reaches the target value Ym1, the scanner driver sets the output value “Yr1” of the point Pr1 based on the difference ΔYrp1 obtained here (S620). In this way, the output value “Yr1” of the point Pr1 is set so that the average value AVr1 of the density values of the pixels occupying the small area Ry1 of the corrected red (R) histogram becomes the target value Ym1. When the output value “Yr1” of the point Pr1 is set to a value smaller than the output value “Yrt1”, it is preferable to check in step S618 whether or not the average value AVr1 is lower than the target value Ym1.

  On the other hand, for the output value “Yr2” of the point Pr2, the same method is used so that the average value AVr2 of the density values of the pixels occupying the small area Ry2 of the corrected red (R) histogram becomes the target value Ym2. The output value “Yr2” of Pr2 is set. A procedure for setting the output value “Yr2” of the point Pr2 is shown in FIG. 31B.

  The scanner driver adjusts tone curves Tcg and Tcb of green (G) and blue (B) other than red (R) based on the two target values Ym1 and Ym2 by such a method.

<Correction process>
The scanner driver then applies an intermediate to the image read by the image reading device 10 based on the tone curves Tcr, Tcg, Tcb for each color in the “density correction” adjusted based on the correction information. A correction process is performed for a backlight image in which the color balance of the tone is lost. Specifically, in the “density correction” in the adjustment procedure described in FIG. 11, the image reading device 10 uses the tone curves Tcr, Tcg, Tcb for each color in the “density correction” adjusted based on the correction information. The density value of each color of the data of each pixel constituting the read image is converted. As a result, the scanner driver performs correction processing for the backlight image in which the halftone color balance is lost with respect to the image read by the image reading device 10.

  Here, the image subjected to the correction process for the backlight image in which the halftone color balance is lost has already been adjusted in accordance with the adjustment procedure shown in FIG. 11 and the histogram adjustment and image adjustment (excluding (3) saturation adjustment). ) Etc. In this way, the scanner driver can process an image of a backlight image whose halftone color balance has been lost by “density correction” on an image that has already undergone histogram adjustment or image adjustment (excluding (3) saturation adjustment). Correction processing is performed.

  32A and 32B show average values of RGB colors before and after correction processing of a backlight image in which the halftone color balance is lost. FIG. 32A shows an average value of each color of RGB before correction processing. FIG. 32B shows an average value of each color of RGB after the correction processing.

  Prior to the correction process, as shown in FIG. 32A, the average value of the density values of the RGB colors of the pixels occupying the three small areas Ry1, Ry2, and Ry3 is the other small areas in the small area Ry2, that is, the halftone area. The variation is larger than that of the regions Ry1 and Ry3.

  On the other hand, after the correction process, as shown in FIG. 32B, the average values AVr1, AVg1, and AVb1 of the RGB small color regions Ry1 are set to be the target values Ym1 common to the respective colors. For this reason, the average values AVr1, AVg1, and AVb1 of the RGB sub-regions Ry1 are all set to substantially the same value. Similarly, the average values AVr2, AVg2, and AVb2 of the RGB small regions Ry2 are set so as to be the target values Ym2 common to the respective colors. For this reason, the average values AVr2, AVg2, and AVb2 of the small areas Ry2 of the RGB colors are all set to substantially the same value.

  FIG. 33 illustrates an example of tone curves Tcr, Tcg, and Tcb for each of the RGB colors generated in this correction process. The tone curves Tcr, Tcg, and Tcb for each color of RGB are formed in different shapes as shown in FIG. This is because arbitrary points set to form tone curves Tcr, Tcg, and Tcb for each color of RGB (for example, points Pr1, Pr2, and Pr3 in the case of the tone curve Tcr of red (R)) This is because each is set individually. In this way, the tone curves Tcr, Tcg, and Tcb for each of the RGB colors are individually formed, so that an appropriate correction process can be performed on the backlight image in which the halftone color balance is lost.

  FIG. 34A and FIG. 34B show an example of a histogram for each color of RGB before and after the correction process. FIG. 34A shows a histogram of each color of RGB before correction processing. FIG. 34B shows a histogram of each color of RGB after the correction process.

  Prior to the correction process, as shown in FIG. 34A, the histograms of red (R) and green (G) are not significantly different and have substantially the same characteristics, whereas the histogram of blue (B) Are significantly different in shape from the red (R) and green (G) histograms and have different characteristics. On the other hand, after the correction process, as shown in FIG. 34B, the blue (B) histogram is improved to have substantially the same characteristics as the red (R) and green (G) histograms.

=== Correction of backlight image of sunset and sunrise ===
Next, “correction processing for a backlight image of sunset or sunrise” will be described. This correction process is performed when it is determined that the image is a sunset or sunrise image in the “determination process for determining whether the image is a sunset or sunrise image”.
This correction process is performed using the histograms for each of the RGB colors generated in the previous “backlight image determination”. That is, as in the case of “correction processing for a normal backlight image” and “correction processing for a backlight image in which the halftone color balance is lost”, each of the RGB colors generated in the previous “backlight image determination” is determined. Attribute information is obtained from each of the three small areas Ry1, Ry2, Ry3 out of the four small areas Ry1, Ry2, Ry3, Ry4 obtained by dividing in the histogram, and an image is obtained based on the obtained attribute information. Correction processing is performed on the image. The correction method will be described in detail below.

<Acquisition of attribute information>
First, the scanner driver performs three small areas Ry1, Ry2, Ry1, Ry3, Ry4 out of the four small areas Ry1, Ry2, Ry3, Ry4 obtained by segmenting in the histogram for each of the RGB colors generated in the above-mentioned “backlight image determination”. , Ry3 respectively to obtain attribute information. Here, as the attribute information, the average values AVr1, AVr2, and AVr3 of the density values of the pixels that occupy the three small regions Ry1, Ry2, and Ry3 from the histogram of each color of RGB are the same as in the case of “normal backlight image correction”. , AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3 are obtained.

<Generation of correction information>
Thereafter, the scanner driver generates correction information for performing correction processing on the backlight image of the sunset or sunrise based on the acquired average values AVr1, AVr2, AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3. . Hereinafter, a correction information generation method according to the present embodiment will be described.

(1) Acquisition of information related to luminance Here, the scanner driver obtains the average obtained in the same manner as in the case of “normal backlight image correction” and “backlight image correction processing in which halftone color balance is lost”. Based on the values AVr1, AVr2, AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, AVb3, the average values Yav1, Yav2 of the luminances of the pixels occupying the respective regions Ry1, Ry2, Ry3 for the three small regions Ry1, Ry2, Ry3, respectively. , Yav3 is acquired. At this time, the scanner driver uses, for example, the above formulas (10) to (12). Further, the scanner driver obtains the average luminance value Yav0 of the entire image read by the image reading device 10 based on the acquired average luminance values Yav1, Yav2, and Yav3. At this time, the scanner driver uses the above equation (13), for example.

(2) Acquisition of Target Value Next, the scanner driver obtains the average luminance values Yav1, Yav2, and Yav3 of the pixels that occupy the three small regions Ry1, Ry2, and Ry3 and the average luminance value Yav0 of the entire image. Based on the above, the target value is obtained. This target value is a value acquired for executing the correction process on the sunset or the backlight image of the sunset, and is a value serving as an index for determining the outline of the correction process. In the present embodiment, two target values Ym1 and Ym2 are acquired as the target values in order to correct a backlight image of sunset or sunrise. Among these, the target value Ym1 is acquired corresponding to the small region Ry1. Further, the target value Ym2 is acquired corresponding to the small region Ry2.

  These two target values Ym1 and Ym2 are the luminance values of the pixels occupying the small region Ry1 as in the case of “normal backlight image correction” or “backlight image correction processing in which the halftone color balance is lost”. Are obtained by the relational expressions (A) and (B) shown in FIG. That is, the target value Ym1 is acquired as a value corresponding to the added value Sadd from the relational expression (A). The target value Ym2 is acquired as a value corresponding to the added value Sadd from the relational expression (B).

(3) Generation of correction information The scanner driver generates correction information based on the two target values Ym1 and Ym2 acquired in this way. This correction information is information for performing an appropriate correction process on a sunset or sunrise backlight image. In the present embodiment, the scanner driver performs the same correction process for a sunset or sunrise backlight image as in the case of “normal backlight image correction” or “backlight image correction process in which halftone color balance is lost”. In addition, the above-described “density correction” is executed. Therefore, in this embodiment, the scanner driver generates setting information for “density correction” as correction information. A method for generating the setting information will be described in detail below.

  The scanner driver performs the “density correction” on the backlight image of the sunset or sunrise, as described in FIG. 8, as in the case of the “correction process of the backlight image in which the halftone color balance is lost”. , Red (R), green (G) and blue (B) tone curves Tcr, Tcg and Tcb are individually adjusted. Here, the scanner driver adjusts the tone curves Tcr, Tcg, and Tcb of each color of red (R), green (G), and blue (B) based on the acquired two target values Ym1 and Ym2. As a method of adjusting the tone curves Tcr, Tcg, and Tcb of each of the RGB colors based on the two target values Ym1 and Ym2, the “correction processing of a backlight image in which the halftone color balance is lost” described in FIG. The same method is adopted.

  However, in the adjustment method of the tone curves Tcr, Tcg, and Tcb for each of the RGB colors performed here, a partial adjustment method is different from the above-described “correction processing of a backlight image in which the halftone color balance is lost”. Is different. That is, in the case of the above-described “correction processing of a backlight image in which the halftone color balance is lost”, the two target values Ym1 and Ym2 are common to each color when adjusting the tone curves Tcr, Tcg, and Tcb of the RGB colors. It was used as a target value. However, in the correction process performed here, among the two target values Ym1 and Ym2, the target value Ym1 is used as a common target value for each color, but the target value Ym2 is red (R), green (G ) And blue (B), not as a common target value for each color, but as a target value for only one of the colors. As the target values for the other colors, different target values obtained individually based on the target value Ym2 are used. That is, a target value for each color is used. Here, the color for which the target value Ym2 is used as it is is set to the color having the largest average value of the density values of the pixels occupying the small region Ry2. For the other colors, target values are individually set based on the target value Ym2 according to the average value of the density values of the pixels occupying the small region Ry2.

  Specifically, a method of adjusting the red (R) tone curve Tcr will be described as an example. As shown in FIG. 30, the scanner driver sets three points Pr1, Pr2, and Pr3 through which the tone curve Tcr passes based on the two target values Ym1 and Ym2 as correction information. The point Pr1 is set corresponding to the small area Ry1, the point Pr2 is set corresponding to the small area Ry2, and the point Pr3 is set corresponding to the small area Ry3. The input values of the points Pr1, Pr2, and Pr3 are “Xr1”, “Xr2”, and “Xr3”, respectively, and the output values of the points Pr1, Pr2, and Pr3 are “Yr1”, “Yr2”, and “Yr3”, respectively. Here, the input values “Xr1”, “Xr2”, and “Xr3” of the points Pr1, Pr2, and Pr3 are the input values “X1”, “X2”, and “X3” of the points Ps1, Ps2, and Ps3 described in FIG. Set in the same way as

  On the other hand, the output values “Yr1” and “Yr2” of the points Pr1 and Pr2 are set based on the two target values Ym1 and Ym2. The output value “Yr3” of the point Pr3 is set on a straight line connecting the point Pr2 and the point Pr0. The point Pr0 is a point where the input value is “255” and the output value is “255”. If the input value “Xr3” of the point Pr0 is determined, the output value “Yr3” can be obtained.

  The output values “Yr1” and “Yr2” of the points Pr1 and Pr2 are obtained by the same method as in the case of “correction of a backlight image in which the halftone color balance is lost”. However, the setting method for the output value “Yr2” of the point Pr2 is different depending on each color of red (R), green (G), and blue (B). That is, for any one of red (R), green (G), and blue (B), the target value Ym2 is used as it is, but the other colors are individually obtained based on the target value Ym2. Different target values are used.

Here, it is assumed that the image to be corrected is a sunset image having histogram characteristics as shown in FIGS. 22A to 22C. In this case, the color having the highest average density value of the pixels occupying the small area Ry2 is red (R). Therefore, the color for which the target value Ym2 is used as it is is set to red (R). On the other hand, for other colors, that is, green (G) and blue (B), target values are individually set based on the average value of the density values of the pixels occupying the small region Ry2 based on the target value Ym2. The If the target value of green (G) set here is “Mg2” and the target value of blue (B) is “Mb2”, these target values “Mg2” and “Mb2” have the following relationship: It can obtain | require by Formula (14) and (15).
Mg2 = Ym2 × AVg2 / AVr2 (14)
Mb2 = Ym2 × AVb2 / AVr2 (15)
These relational expressions (14) and (15) are for setting the ratio of the average value of the density values of the pixels occupying the small area Ry2 of each RGB color after correction to be substantially the same as the ratio before correction. Is. If the target values “Mg2” and “Mb2” are obtained using the relational expressions (14) and (15), the color balance before correction can be maintained in the corrected halftone area.

  Thereby, for red (R), the output value “Yr2” of the point Pr2 is set so that the average value AVr2 of the density values of the pixels occupying the small region Ry2 of the corrected red (R) histogram becomes the target value Ym2. Is set.

  On the other hand, green (G) and blue (B) are set as follows. FIG. 35 illustrates a method for adjusting the green (G) tone curve Tcg. FIG. 36 illustrates a method for adjusting the blue (B) tone curve Tcb. The points for adjusting the green (G) tone curve Tcg are Pg1, Pg2, and Pg3 as shown in FIG. 35, and the points for adjusting the blue (B) tone curve Tcb are shown in FIG. Thus, Pb1, Pb2, and Pb3 are used. The output value of the green (G) point Pg1 is “Yg1”, the output value of the point Pg2 is “Yg2”, and the output value of the point Pg3 is “Yg3”. Further, the output value of the point Pb1 of blue (B) is “Yb1”, the output value of the point Pb2 is “Yb2”, and the output value of the point Pb3 is “Yb3”.

  For the green (G) point Pg2, the output value “Yg2” of the point Pg2 is set so that the average value AVg2 of the density values of the pixels occupying the small region Ry2 of the corrected green (G) histogram becomes the target value Mg2. Is set. For the blue (B) point Pb2, the output value “Pb2” is set so that the average value AVb2 of the density values of the pixels occupying the small region Ry2 of the corrected blue (B) histogram becomes the target value Mb2. Yb2 ”is set.

  For the output values “Yr1”, “Yg1”, and “Yb1” of the points Pr1, Pg1, and Pb1, the target value Ym1 is used as a common target value for each color of RGB. That is, the average values AVr1, AVg1, and AVb1 of the density values of the pixels that occupy the small area Ry1 of the histogram of each color of red (R), green (G), and blue (B) after correction become the target value Ym1. , RGB output values “Yr1”, “Yg1”, “Yb1” of the points Pr1, Pg1, and Pb1 are set.

  As a method for obtaining the output values “Yr1”, “Yg1”, “Yb1” of the points Pr1, Pg1, and Pb1 of the tone curves Tcr, Tcg, and Tcb of each color of red (R), green (G), and blue (B) The method described in FIG. 31A can be used. Further, as a method of obtaining the output value “Yr2” of the point Pr2 of the red (R) tone curve Tcr, the method described in FIG. 31B can be used.

  On the other hand, the following methods are used to obtain the output values “Yg2” and “Yb2” of the points Pg2 and Pb2 of the tone curves Tcg and Tcb of green (G) and blue (B). FIG. 37A shows an example of a method for acquiring the output value “Yg2” of the point Pg2 of the tone curve Tcg of green (G). FIG. 37B shows an example of a method for acquiring the output value “Yb2” of the point Pb2 of the blue (B) tone curve Tcb. In the methods shown in FIGS. 37A and 37B, the output value “Yg2” of the point Pg2 and the output value “Yb2” of the point Pb2 are obtained by the same method as the method shown in FIGS. 31A and 31B described above. get. However, the output value “Yg2” of the green (G) point Pg2 is set based on the target value Mg2, as shown in step S718 in FIG. 37A. Further, the output value “Yb2” of the blue (B) point Pb2 is set based on the target value Mb2, as shown in step S768 in FIG. 37B.

<Correction process>
Then, the scanner driver generates three points Pr1, Pr2, Pr3, Pg1, Pg2, Pg3 for each color for adjusting the tone curves Tcr, Tcg, Tcb of each color in “density correction” generated as correction information in this way. Based on the information on Pb1, Pb2, and Pb3, the image read by the image reading device 10 is subjected to correction processing for a backlight image of sunset or sunrise. Specifically, in the “density correction” in the adjustment procedure described with reference to FIG. 11, information regarding three points Pr1, Pr2, Pr3, Pg1, Pg2, Pg3, Pb1, Pb2, and Pb3 generated as correction information is included. Based on the tone curves Tcr, Tcg, and Tcb adjusted based on the above, the density value of each color of the data of each pixel constituting the image read by the image reading device 10 is converted. As a result, the scanner driver performs a correction process for the backlight image of sunset or sunrise on the image read by the image reading device 10.

  Here, the image subjected to the correction process for the sunset or sunrise backlight image has already been subjected to histogram adjustment, image adjustment (excluding (3) saturation adjustment) and the like according to the adjustment procedure shown in FIG. It is an image that was made. The scanner driver performs correction processing for a backlight image at sunset or sunrise by “density correction” on an image that has already been subjected to histogram adjustment or image adjustment (excluding (3) saturation adjustment). Apply.

  FIG. 38A and FIG. 38B show average values of RGB colors before and after correction processing of a sunset or morning backlight image. FIG. 38A shows an average value of each color of RGB before the correction process. FIG. 38B shows the average value of each RGB color after the correction processing.

  Before the correction processing, as shown in FIG. 38A, the average value of the density values of the RGB colors of the pixels occupying the three small areas Ry1, Ry2, and Ry3 is the specific color in the small area Ry3, that is, the highlight area. (Red (R)) is larger. On the other hand, in the small areas Ry1, Ry2, that is, in the shadow area and the halftone area, the average value of the density values of the RGB colors is very small.

  On the other hand, after the correction processing, as shown in FIG. 38B, the average values AVr1, AVg1, and AVb1 of the RGB small color regions Ry1 are set to be the target values Ym1 common to the respective colors. For this reason, the average values AVr1, AVg1, and AVb1 of the RGB small color regions Ry1 are all set to substantially the same value, and the image quality of the shadow region is improved.

  In addition, for the RGB small area Ry2, the average values AVr2, AVg2, and AVb2 of the RGB small areas Ry2 are set to individual target values Ym2 so that the ratio of the average values of the RGB colors before correction is maintained. , Mg2, and Mb2. Therefore, the average values AVr2, AVg2, and AVb2 of the RGB small color regions Ry2 improve the image quality of the halftone region while maintaining the ratio before correction.

  FIG. 39 illustrates an example of tone curves Tcr, Tcg, and Tcb for each color of RGB generated in this correction process. The tone curves Tcr, Tcg, and Tcb for each color of RGB are formed in different shapes as shown in FIG. This is because arbitrary points set to form tone curves Tcr, Tcg, and Tcb for each color of RGB (for example, points Pr1, Pr2, and Pr3 in the case of the tone curve Tcr of red (R)) This is because each is set individually. In this case, as shown in the figure, the tone curves Tcr, Tcg, Tcb for each of the RGB colors are applied so that the halftone area and the highlight area are sufficiently corrected in addition to the shadow area. Is formed. As described above, the tone curves Tcr, Tcg, and Tcb for each color of RGB are formed, so that an appropriate correction process can be performed on the backlight image of sunset or sunrise.

  FIG. 40A and FIG. 40B show an example of a histogram for each color of RGB before and after the correction process. FIG. 40A shows a histogram of each color of RGB before correction processing. FIG. 40B shows a histogram of each color of RGB after the correction process.

  Before the correction processing, as shown in FIG. 40A, the image has a typical backlight image characteristic in which pixels are biased between a shadow area and a highlight area. On the other hand, after the correction processing, as shown in FIG. 40B, the histogram of each color of red (R), green (G), and blue (B) is entirely from the shadow region to the halftone region and the highlight region. It has been improved to become brighter.

=== Other application examples ===
Here, the regions represented by the histograms of three colors of red (R), green (G), and blue (B) are divided into four small regions Ry1, Ry2, Ry3, and Ry4. It is not necessary to divide into four small regions Ry1, Ry2, Ry3, Ry4. That is, the area represented by the histogram of three colors of red (R), green (G) and blue (B) may be divided into at least three or more small areas. It may also be divided into five or more small areas.

  In addition, here, four small regions Ry1, Ry2, Ry3, and Ry4 obtained by dividing the region represented by the histogram of three colors of red (R), green (G), and blue (B) Included three small regions Ry1, Ry2, and Ry3 having almost the same area and the same number of pixels. However, it is necessary to provide a small region having substantially the same area and the same number of pixels. Absent. In addition, the present invention is not necessarily limited to the case where three small regions having substantially the same area or number of pixels are provided as described above, and may be two or four or more.

  Further, here, in the “correction processing of a backlight image in which the halftone color balance is lost”, the regions represented by the histograms of three colors of red (R), green (G), and blue (B) are classified. In order to acquire attribute information from the obtained four small areas Ry1, Ry2, Ry3, and Ry4, three small areas Ry1, Ry2, and Ry3 have been selected. It is not limited to. That is, at least one small region is selected from two or more small regions obtained by dividing the region represented by the histogram of three colors of red (R), green (G), and blue (B). It suffices to select one of the three small regions Ry1, Ry2, and Ry3. In addition, when the small area is divided by a method other than the method described here, the small area may be selected by another method.

  Further, here, in the “correction processing of a backlight image in which the halftone color balance is lost”, three small regions Ry1, Ry2, and Ry3 are occupied for each color as attribute information from the three small regions Ry1, Ry2, and Ry3. Although the average values AVr1, AVr2, AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3 of the density values of the pixels are acquired, such attribute information is not necessarily acquired. In other words, the attribute information acquired from the selected small area may be any attribute information, for example, the upper limit value or lower limit value of the density value of the pixels occupying the selected small areas Ry1, Ry2, Ry3, It may be the size of the area of the selected small regions Ry1, Ry2, and Ry3. Note that the attribute information acquired from the selected small area is preferably information that is useful in performing “correction processing of a backlight image in which the halftone color balance is lost”.

  In addition, here, in the “correction process for a backlight image in which the halftone color balance is lost”, the pixels occupying the small regions Ry1, Ry2, and Ry3 for each color obtained as attribute information from the small regions Ry1, Ry2, and Ry3, respectively. Based on the average values AVr1, AVr2, AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3 of the density value, information on luminance is obtained, and two target values Ym1 and Ym2 are obtained based on the information on the luminance. Although the correction processing is executed based on the target values Ym1 and Ym2, the correction processing performed here does not necessarily have to be executed by such a method. That is, the target value may be acquired based on the attribute information acquired from the three small areas Ry1, Ry2, and Ry3, and the correction process may be executed by another method.

  Further, here, the “correction processing of the backlight image in which the halftone color balance is lost” is performed by adjusting the tone curves Tcr, Tcg, Tcb in “density correction”, but such a method is not necessarily performed. Therefore, the correction process need not be performed. That is, any correction process may be executed as long as it is a correction process for a backlight image whose halftone color balance is lost.

=== Summary ===
As described above, in the present embodiment, a histogram of three colors of red (R), green (G), and blue (B) is generated based on the data of each pixel constituting the image read by the image reading device 10. Then, the regions represented by these histograms are divided into four small regions Ry1, Ry2, Ry3, Ry4, respectively, and three small regions Ry1, Ry2, Ry3, Ry4 are selected from these four small regions Ry1, Ry2, Ry3, Ry4. Ry2 and Ry3 are selected for each color, attribute information is acquired from each of these three small regions Ry1, Ry2, and Ry3, and target values Ym1 and Ym2 are acquired based on these attribute information, and the target value Ym1 , Ym2 is applied to the image read by the image reading device 10 so that “a correction process for a backlight image in which the halftone color balance is lost” is performed. Image read by 0, if the color balance of intermediate tones is an image destroyed, it can be subjected to appropriate correction processing.

  In addition, in “correction processing of backlight image in which halftone color balance is lost”, the regions represented by the histograms of the colors of red (R), green (G), and blue (B) were obtained respectively. In the four small regions Ry1, Ry2, Ry3, Ry4, there are provided three small regions Ry1, Ry2, Ry3 having substantially the same area and the same number of pixels, and attributes from these three small regions Ry1, Ry2, Ry3, respectively. By acquiring the information, when the image read by the image reading device 10 is an image in which the halftone color balance is lost, an appropriate correction process can be performed.

  Further, in the “correction processing of a backlight image in which the halftone color balance is lost”, as the attribute information, the average values AVr1, AVr2, and the average values AVr1, AVr2, of the color values of the pixels occupying the three selected small regions Ry1, Ry2, Ry3, By acquiring AVr3, AVg1, AVg2, AVg3, AVb1, AVb2, and AVb3, a more appropriate correction process is performed when the image read by the image reading device 10 is an image in which the halftone color balance is lost. Can be applied.

  In the “correction process for a backlight image in which the halftone color balance is lost”, the average values AVr1, AVr2, AVr3, AVg1, and the average density values of the colors of the pixels occupying the three selected small regions Ry1, Ry2, Ry3, Based on AVg2, AVg3, AVb1, AVb2, and AVb3, the average luminance values Yav1, Yav2, and Yav3 of the pixels that occupy the three small regions Ry1, Ry2, and Ry3 are obtained, so that the image read by the image reading device 10 can be obtained. In the case of an image in which the halftone color balance is lost, more appropriate correction processing can be performed.

  In addition, in the “correction processing of a backlight image in which a halftone color balance is lost”, the average luminance value Yav1 of the pixels occupying the acquired small region Ry1 and the average luminance value Yav0 of the entire image are added and added. By acquiring the value Sadd, a more appropriate target value can be acquired. Accordingly, when the image read by the image reading device 10 is an image in which the halftone color balance is lost, more appropriate correction processing can be performed.

  In addition, by executing “density correction” as “backlight image correction processing in which the halftone color balance is lost”, an image read by the image reading apparatus 10 can be easily displayed. A correction process for a broken backlight image ”can be performed.

=== Application to printing apparatus ===
Such an image reading apparatus 10 may be mounted on a printing apparatus. FIG. 41 illustrates a configuration example of the printing apparatus 220 including such an image reading apparatus 10. The printing apparatus 220 includes a scanner function for reading an image from a document and generating image data, a printer function for printing on various media such as printing paper based on print data sent from the host computer 248, and a reading function from the document. And a local copy function for copying a printed image on a medium. The printing apparatus 220 includes a scanner unit 222 that reads an image from a document (including the scanner control unit 238 and corresponds to the “image reading apparatus 10” herein), and a printer unit 224 that performs printing on a medium S such as printing paper. And.

  Further, the control unit 226 of the printing apparatus 220 includes a CPU 228, a memory 230, an external communication interface 232, an operation input interface 234, a display control unit 236, a scanner control unit 238, as shown in FIG. An image processing unit 240, a printer control unit 242, a card interface 244, and a bus 246 for connecting them to each other are provided.

  The CPU 228 controls the entire printing apparatus 220. The memory 230 stores a program executed by the CPU 228, data used in the program, and the like. The external communication interface 232 communicates with the host computer 248 by wire or wireless. The operation input interface 234 receives an operation input from the user through the operation button 250 or the like. The display control unit 236 controls a display unit such as the liquid crystal display 252.

  On the other hand, the scanner control unit 238 controls the scanner unit 222 to execute a reading operation for reading an image from a document. The image processing unit 240 executes print processing on the image data output from the scanner unit 222 by using the printer unit 224 in order to print an image read from the document by the scanner unit 222 on a medium by the printer unit 224. To play a role in converting data. The printer control unit 242 controls the printer unit 224. The card interface 244 performs processing for reading image data stored in the memory card from various memory cards set in the card reader unit 254.

  In the scanner function, the printing apparatus 220 outputs image data read from the document by the scanner unit 222 to the host computer 248. In the printer function, the printing apparatus 220 performs printing on various media by the printer unit 224 based on the print data sent from the host computer 248. In the local copy function, the printing apparatus 220 prints an image read from a document by the scanner unit 222 on various media by the printer unit 224 and copies it.

=== Other Embodiments ===
The present invention has been described above by taking an embodiment of an image reading system as an example. However, the above embodiment is for facilitating the understanding of the present invention, and is intended to limit the present invention. Is not. The present invention can be changed or improved without departing from the gist thereof, and needless to say, the present invention includes equivalents thereof. In particular, the embodiments described below are also included in the present invention.

<About images>
In the above-described embodiment, the image read by the image reading apparatus 10 has been described as an example. However, the “image” referred to here may be any image. For example, specifically, an image taken by a digital camera or the like may be used. The image format may be any type, for example, an image expressed in various formats such as a JPEG format, a bitmap format, and a YUV format.

<About concentration values>
In the above-described embodiment, the density value expressed in 256 gradations, that is, the density value that can take values of “0” to “255”, for example, has been described as an example. The “density value” is not limited to such a density value. In other words, the “density value” here may be any density value as long as it is data for expressing the density of each pixel constituting the image. The “density value” includes a brightness density value expressed in a YUV format or the like.

<About the histogram>
In the embodiment described above, histogram data of three colors of red (R), green (G), and blue (B) is generated as “histogram data”. In this case, it is not always the case that the histogram data of these three colors of red (R), green (G) and blue (B) is generated. That is, at least two types of histogram data of different colors may be generated as “histogram data”.

<About the histogram data generator>
In the above-described embodiment, as the “histogram data generation unit”, the scanner driver generates histogram data based on the data of the pixels constituting the image, that is, the image read by the image reading device 10 here. However, the “histogram data generation unit” here does not necessarily need to be such a scanner driver. In other words, in the “histogram data generation unit” here, any type of “histogram data generation unit” can be used as long as histogram data is generated based on the data of the pixels constituting the image to be determined. It does not matter.

<About the attribute information acquisition unit>
In the above-described embodiment, as the “attribute information acquisition unit”, the scanner driver determines that the region represented by the histogram of each color of red (R), green (G), and blue (B) is divided into three by density value size. It was divided into two or more small areas, and at least one small area was selected from these three or more small areas, and attribute information was acquired from the selected small areas. The “information acquisition unit” does not necessarily need to be such a scanner driver. In other words, the “attribute information acquisition unit” here may be any type of “attribute information acquisition unit” as long as the attribute information is acquired by the method described above.

<About the target value acquisition unit>
In the above-described embodiment, as the “target value acquisition unit”, the scanner driver classifies the region represented by the histogram of each color of red (R), green (G), and blue (B) according to the size of the density value. Acquired by selecting at least one small area with a density value larger than a small area and smaller density value than other small areas for each color. Based on the attribute information, the target value common to each color was acquired for the attribute information related to the selected small area, but the “target value acquisition unit” here does not necessarily need to be such a scanner driver. . In other words, the “target value acquisition unit” here is not limited to the scanner driver described above, and may be of any type.

<Image correction device>
In the above-described embodiment, the “image correction apparatus” has been described by taking an example of an apparatus that performs correction processing on a backlight image in which color balance such as halftone is lost for an image read by the image reading apparatus 10. The “image correction apparatus” here is not limited to an apparatus that performs correction processing on an image read by the image reading apparatus 10. That is, the type of image is not limited, and any type of device is included in the “image correction device” as long as the device performs correction processing on various images.

<About image correction program>
In the above-described embodiment, the scanner driver is executed as the “image correction program” in the computer device 20 that is communicably connected to the image reading device 10 by wired or wireless communication. Is not limited to such a scanner driver. In other words, the “image correction program” here may be any program as long as it is a program that performs correction processing on a backlight image in which the color balance such as halftone is lost.

<About image reader>
In the above-described embodiment, the “image reading device” is an image reading device in which the image sensor 50 that reads an image is provided on the carriage 40 that moves relative to the document 15 set on the document table 12. As described above, the “image reading apparatus” here is not necessarily limited to this type of “image reading apparatus”. In other words, any type of apparatus is included in the “image reading apparatus” as long as the apparatus reads an image from a document. That is, the “image reading device” here includes an image reading device in which the document 15 moves relative to the image sensor 50 that reads an image and reads an image from the document 15.

<About printing devices>
In the above-described embodiment, as the “printing device”, an image reading unit (scanner unit 222) that reads an image from the document 15 and generates image data, and a printing unit (printer unit 224) that performs printing on a medium. However, the “printing apparatus” described here is not necessarily limited to the image reading unit (scanner unit 222 (scanner control unit 238 or image processing unit 240). Etc.))) is not necessary. In other words, the “printing apparatus” here may include at least a printing unit that performs printing on a medium.

1 is an explanatory diagram of an embodiment of an image reading system. FIG. FIG. 3 is an explanatory diagram illustrating an example of an internal configuration of an image reading apparatus. 1 is an explanatory diagram illustrating an example of a system configuration of an image reading apparatus. FIG. Explanatory drawing of an example of the system configuration | structure of a computer main body. FIG. 3 is an explanatory diagram illustrating an example of a main dialog box of a scanner driver. FIG. 6 is an explanatory diagram illustrating an example of a dialog box for adjusting a histogram. Explanatory drawing about the definition method of a tone curve. FIG. 4 is an explanatory diagram of an example of a density correction dialog box. Explanatory drawing of an example of the dialog box of an image adjustment. Explanatory drawing of the data set by histogram adjustment. Explanatory drawing of the data set by density correction. Explanatory drawing of the data set by image adjustment. Explanatory drawing of an example of the adjustment procedure of an image. Explanatory drawing of an example of an image reading procedure. 6 is a flowchart illustrating an image correction method according to the present embodiment. 6 is a flowchart for explaining an example of a processing procedure for backlight correction of a scanner driver. The flowchart explaining the determination method of a backlight image. An explanatory view of an example of a histogram. Explanatory drawing in the case of determining not to be a backlight image. Explanatory drawing in the case of determining with it being a backlight image. FIG. 6 is an explanatory diagram illustrating an example of a red (R) histogram of an image in which a halftone color balance is lost. FIG. 5 is an explanatory diagram illustrating an example of a green (G) histogram of an image in which a halftone color balance is lost. Explanatory drawing of an example of the histogram of the blue (B) of the image where the color balance of the halftone was broken. Explanatory drawing of the selection method of the small area | region in the determination of the image in which the color balance of halftone collapsed. Explanatory drawing when it determines with it being an image in which the color balance of the halftone collapsed. Explanatory drawing in the case of determining that it is not an image in which the halftone color balance is lost. 6 is a flowchart illustrating an example of an image determination process in which a halftone color balance is lost. Explanatory drawing of an example of the histogram of red (R) of the image of sunset or sunrise. Explanatory drawing of an example of the histogram of green (G) of the image of sunset or sunrise. Explanatory drawing of an example of the histogram of blue (B) of the image of sunset or sunrise. Explanatory drawing of the selection method of the small area | region in the determination of the image of sunset / sunrise. Explanatory drawing in the case of determining with it being an image of sunset or sunrise. Explanatory drawing when it determines with it not being an image of a sunset or a morning glow. The flowchart which shows an example of the determination process of the image of sunset or sunrise. Explanatory drawing of an example of the acquisition method of target value. Explanatory drawing of an example of the production | generation method of correction information. Explanatory drawing of an example of the setting method of point Ps1. Explanatory drawing of an example of the setting method of point Ps2. Explanatory drawing of an example of the other setting method of input value "X1" of point Ps1. Explanatory drawing of an example of the other setting method of input value "X2" of point Ps2. Explanatory drawing of an example of the other setting method of input value "X3" of point Ps3. Explanatory drawing of an example of the adjustment method of the tone curve of red (R). The flowchart explaining an example of the setting procedure of point Pr1. The flowchart explaining an example of the setting procedure of point Pr2. Explanatory drawing of the average value of each RGB color before the correction process to the backlight image where the halftone color balance collapsed. Explanatory drawing of the average value of each RGB color after the correction process to the backlight image in which the color balance of the halftone collapsed. Explanatory drawing of an example of the tone curve of each color of RGB. Explanatory drawing of an example of the histogram of each RGB color before a correction process. Explanatory drawing of an example of the histogram of each RGB color after a correction process. Explanatory drawing of an example of the adjustment method of the tone curve of green (G). Explanatory drawing of an example of the adjustment method of the tone curve of blue (B). The flowchart explaining an example of the setting procedure of point Pg2. The flowchart explaining an example of the setting procedure of the point Pb2. Explanatory drawing of the average value of each RGB color before the correction process to the backlight image of sunset or morning glow. Explanatory drawing of the average value of each RGB color after the correction process to the backlight image of sunset or morning glow. Explanatory drawing of an example of the tone curve of each color of RGB. Explanatory drawing of an example of the histogram of each RGB color before a correction process. Explanatory drawing of an example of the histogram of each RGB color after a correction process. FIG. 3 is an explanatory diagram illustrating an embodiment of a printing apparatus.

Explanation of symbols

2 image reading system, 10 image reading device, 12 document table,
14 Document cover, 15 Document, 18 Hinge, 20 Computer device,
22 computer body, 24 display device, 26 input device,
28 FD drive device, 30 CD-ROM drive device,
32 readers, 34 keyboards, 36 mice, 40 carriages,
42 Drive mechanism, 44 guide, 46 exposure lamp, 48 lens,
50 image sensor, 52 control unit, 54 timing belt,
55 pulley, 56 pulley, 58 drive motor, 60 controller,
62 motor control unit, 64 lamp control unit, 66 sensor control unit,
68 AFE (Analog Front End) part, 70 Digital processing circuit,
72 interface circuit, 74 analog signal processing circuit,
76 A / D conversion circuit, 80 CPU, 82 memory,
84 HDD (Hard Disk Drive Device), 86 Operation Input Unit,
88 display control unit, 90 external communication unit, 92 bus,
100 dialog box, 102 “mode selection field”,
104 “Save settings”, 106 “Original setting field”, 108 “Output setting field”,
110 “Adjustment field”, 112 “Document type”, 114 “Scanning device”,
116 “Auto Exposure”, 118 “Image Type”, 120 “Resolution”,
122 “Original Size”, 124 “Output Size”,
126 “Scan Button”, 127 “Preview Button”,
128A button, 128B button, 128C button, 128D button,
130A check box, 130B check box,
130C check box, 130D check box,
130E check box, 131 dialog box,
132 Histogram display field, 134 Tone curve display field,
136 Gray balance adjustment field, 138 channel field,
140A slider, 140B slider,
140C slider, 142A numeric input field, 142B numeric input field,
142C Numerical input field, 142D Numerical input field, 142E Numerical input field,
143A syringe button, 143B syringe button,
143C dropper button, 144A end curve shape change button,
144B End curve shape change button, 145 slider,
150 dialog box, 152 tone curve display area,
154 Channel field, 156A Numerical value input field, 156B Numerical value input field,
158 Density correction setting save field, 160 dialog box,
162A slider, 162B slider, 162C slider,
162D slider, 162E slider, 162F slider,
164A Numerical input field, 164B Numerical input field, 164C Numerical input field,
164D numeric input field, 164E numeric input field, 164F numeric input field,
220 printing device, 222 scanner unit, 224 printer unit,
226 control unit, 228 CPU, 230 memory,
232 External communication interface, 234 operation input interface,
236 Display control unit, 238 Scanner control unit, 240 Image processing unit,
242 Printer control unit, 244 card interface,
246 bus, 248 host computer, 250 operation buttons,
252 LCD display, 254 card reader

Claims (20)

(A) a histogram data generation unit that generates histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixels based on data of pixels constituting the image;
(B) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small region from the three or more small regions and acquires attribute information about the small region for each color;
(C) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small region, based on the attribute information acquired by the attribute information acquisition unit;
(D) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small area becomes a target value common to the colors acquired by the target value acquisition unit;
An image correction apparatus comprising (E).   The image correction apparatus according to claim 1, wherein the histogram data generation unit generates histogram data for each color of red, green, and blue as the histogram data for each color.   The image correction apparatus according to claim 1, wherein the three or more small regions include two or more small regions having the same area.   The image correction apparatus according to claim 1, wherein the three or more small regions include two or more small regions having the same number of pixels. The three or more small regions include two or more small regions in which at least one of the number and area of the pixels is equal to each other,
The attribute information acquisition unit selects at least one small region from two or more small regions in which at least one of the number and area of the pixels is equal to each other, and acquires the attribute information regarding the small region And
The said target value acquisition part selects the said selection small area | region from two or more small area | regions in which at least any one of the number and the area of the said pixel is mutually equal. The image correction apparatus according to claim 1.   The image correction apparatus according to claim 1, wherein the selected small area is located in a halftone area of the histogram.   The image according to any one of claims 1 to 6, wherein the attribute information acquisition unit acquires, as the attribute information, an average value of the density values for each of the at least one small region for each color. Correction device.   The target value acquisition unit acquires information on luminance for each of the at least one small region based on the average value of the density values for each color acquired by the attribute information acquisition unit, and based on the information on the luminance The image correction apparatus according to claim 7, wherein the target value is acquired.   The image correction apparatus according to claim 8, wherein the target value acquisition unit acquires an average value of luminance of pixels for the at least one small region as the information related to the luminance.   The target value acquisition unit is based on a value obtained by adding at least one of the average brightness values acquired for the at least one small region and the average brightness value of the entire image. The image correction apparatus according to claim 9, wherein the target value is acquired.   The said correction process part converts the density value of each pixel which comprises the said image, when performing the said correction process with respect to the said image, The any one of Claims 1-10 characterized by the above-mentioned. Image correction device.   12. The information processing apparatus according to claim 11, wherein the correction processing unit acquires information about a correspondence relationship between a density value before conversion and a density value after conversion in order to perform the correction process on the image. Image correction device.   The image correction apparatus according to claim 1, further comprising: a determination unit that determines whether or not the image is an image in which a color balance in the selected small area is lost.   The said correction process part performs the said correction process with respect to the said image, when it determines with the color balance in the said selection small area | region having collapsed about the said image by the said determination part. 14. The image correction apparatus according to 13. (A) a histogram data generation unit that generates histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixels based on data of pixels constituting the image;
(B) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small region from the three or more small regions and acquires attribute information about the small region for each color;
(C) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small region, based on the attribute information acquired by the attribute information acquisition unit;
(D) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small area becomes a target value common to the colors acquired by the target value acquisition unit;
(E)
(F) As the histogram data for each color, histogram data for each color of red, green, and blue is generated by the histogram data generation unit,
(G) The three or more subregions include two or more subregions having the same area.
(H) The three or more subregions include two or more subregions having the same number of pixels.
(I) The three or more small regions include two or more small regions in which at least one of the number and area of the pixels is equal to each other, and the attribute information acquisition unit Selecting at least one small region from two or more small regions having at least one of a number and an area equal to each other, and acquiring the attribute information related to the small region; Selecting the selected subregion from two or more subregions in which at least one of the number and area of pixels is equal to each other;
(J) The attribute information acquisition unit acquires, as the attribute information, an average value of the density values for each color for each of the at least one small region,
(K) the selected small area is located in a halftone area of the histogram;
(L) The attribute information acquisition unit acquires, as the attribute information, an average value of the density values for each color for each of the at least one small region,
(M) The target value acquisition unit acquires information on luminance for each of the at least one small region based on an average value of the density values for each color acquired by the attribute information acquisition unit, and relates to the luminance Obtaining the target value based on the information,
(N) The target value acquisition unit acquires an average value of luminance of pixels for the at least one small region as information on the luminance,
(O) The target value acquisition unit is obtained by adding at least one of the average luminance values acquired for the at least one small region and the average luminance value of the entire image. Based on the value, obtain the target value,
(P) When performing the correction process on the image, the correction processing unit converts a density value of each pixel constituting the image,
(Q) In order to perform the correction process on the image, the correction processing unit obtains information regarding a correspondence relationship between a density value before conversion and a density value after conversion,
(R) a determination unit for determining whether or not the image is an image in which the color balance in the selected small area is lost,
(S) The correction processing unit performs the correction processing on the image when the determination unit determines that the color balance of the selected small region is lost for the image. An image correction device. (A) generating histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel based on the data of the pixels constituting the image;
(B) Based on the generated histogram data for each color, dividing the region represented by the histogram for each color into three or more small regions for each density value;
(C) selecting at least one small area from among the three or more small areas and acquiring attribute information relating to the small area for each color;
(D) From among the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. Steps,
(E) acquiring a common target value for each color for the attribute information related to the selected small area based on the acquired attribute information;
(F) performing a correction process on the image so that the attribute information of each color related to the selected small area becomes a target value common to the acquired colors;
An image correction method comprising (G). (A) An image correction program executed by the image correction apparatus,
(B) generating histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel based on the data of the pixels constituting the image;
(C) dividing the region represented by the histogram for each color into three or more small regions according to the density values based on the generated histogram data for each color;
(D) selecting at least one small area from the three or more small areas, and obtaining attribute information about the small area for each color;
(E) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. Steps,
(F) acquiring a common target value for each color for the attribute information related to the selected small area based on the acquired attribute information;
(G) performing a correction process on the image so that the attribute information of each color relating to the selected small area becomes a target value common to the acquired colors;
An image correction program that executes (H). (A) an image reading unit for reading an image from a document;
(B) Histogram data generation that generates histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel based on the data of the pixels constituting the image read by the image reading unit And
(C) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small area from the three or more small areas and acquires attribute information about the small area for each color;
(D) From among the three or more small areas, at least one small area having a density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small area, based on the attribute information acquired by the attribute information acquisition unit;
(E) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small region becomes a target value common to the colors acquired by the target value acquisition unit;
An image reading apparatus comprising (F). (A) a step of reading an image from a document;
(B) generating histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel based on the data of the pixels constituting the read image;
(C) dividing the region represented by the histogram for each color into three or more small regions according to the density values based on the generated histogram data for each color;
(D) selecting at least one small area from the three or more small areas, and obtaining attribute information about the small area for each color;
(E) Out of the three or more small areas, at least one small area having the density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. Steps,
(F) acquiring a common target value for each color for the attribute information related to the selected small area based on the acquired attribute information;
(G) performing a correction process on the image so that the attribute information of each color relating to the selected small area becomes a target value common to the acquired colors;
An image reading method comprising (H). (A) a printing unit that prints an image on a medium;
(B) Before the image is printed by the printing unit, based on the data of the pixels constituting the image, histogram data for each color representing the distribution of the number of pixels with respect to the density value of each color of the pixel is obtained. A histogram data generation unit to generate,
(C) Based on the histogram data for each color generated by the histogram data generation unit, the area represented by the histogram for each color is divided into three or more small areas for each density value. An attribute information acquisition unit that selects at least one small area from the three or more small areas and acquires attribute information about the small area for each color;
(D) From among the three or more small areas, at least one small area having a density value larger than a certain small area and smaller than the other small areas is selected as a selected small area for each color. A target value acquisition unit that acquires a target value common to each color for the attribute information related to the selected small area, based on the attribute information acquired by the attribute information acquisition unit;
(E) a correction processing unit that performs correction processing on the image so that the attribute information of each color related to the selected small region becomes a target value common to the colors acquired by the target value acquisition unit;
A printing apparatus comprising (F).