JP5375675B2 - Image processing apparatus, image processing method, program, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce capacity tightness of an image storage device by estimating contents of various image processing to which a plurality of pieces of image data have been subjected, on the basis of image contents, machine use conditions, etc., thereby obtaining an image conversion method and managing the plurality of pieces of image data only with one template, difference data, and the conversion method. <P>SOLUTION: In order to reduce capacity of an HDD 105, a CPU 106 divides stored images A and B into regions to calculate edge images. The CPU performs first matching processing of the edge images between the image regions and performs second matching processing of color information with respect to matched image regions and registers matched common image regions. With respect to image regions not matched in the first matching processing, the CPU refers to use history information of the machine to estimate an image conversion method, converts one image, and performs the first matching processing again. In the case of not matching, difference information and the conversion method are registered. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to an image processing apparatus, an image processing method, a program, and a recording medium that reduce image capacity compression.

  Normally, an MFP (multifunction peripheral) has a built-in storage device such as an HDD, and image data can be stored. However, the image data has a larger capacity than a text file or the like, and if a large amount of original images is stored, the storage device Squeeze capacity.

  Therefore, as a method of avoiding such a shortage of capacity, when all history information of a document is stored by revision (version) management or the like, instead of storing all history image data, one image data and its There is a method of managing only difference information from image data (see Patent Document 1).

However, since the above-described method manages history information of the same original document as a difference, it cannot be applied to images having the same format and different contents (for example, PPT templates, documents of a predetermined format, etc.). Even when applied, there are the following problems. That is,
1. Even if the format is the same, images stored in the MFP, such as full-color images, two-color images, monochrome images, and negative-positive images, are different. There are various types of originals that support color universal (CUD).

In the case described above, edge information that is not related to color can be matched with a template, but color information must be secured in a separate plane, and there is a problem that the capacity for the plane increases.
2. Aggregated images, scaled images, partially trimmed images (partially copied images), images with punched holes, etc. do not match in matching, so there is a problem that multiple templates are required and the capacity increases. .

The present invention has been made in view of the above problems,
It is an object of the present invention to obtain a method for converting image data by estimating the content of image processing performed on a plurality of image data subjected to various processes from the contents of the image and the usage status of a machine. An object of the present invention is to provide an image processing apparatus, an image processing method, a program, and a recording medium that reduce the capacity compression of the image storage device by managing only with a template, difference data, and a conversion method.

  The present invention estimates the content of image processing performed on a plurality of image data subjected to various processes from the contents of the image and the usage status of the machine, and obtains a method for converting the image data, thereby obtaining one template. The main point is to store only the difference data and the conversion method (without a plurality of templates).

  An outline of the present invention will be described (see FIG. 1). First, image area division is performed on an image stored in an HDD or the like (S1). Next, pattern matching is performed using edge components to determine whether or not the divided template is the same template (S2). If the matching does not match in the edge information, estimation is performed based on the machine usage history information and the image arrangement information, the image is converted, and the matching process is tried again (S4). Even if rematching is performed from S4 to S2, the position coordinates of the image area itself and the image area relative to the image frame are accumulated for an image area in which the number of mismatches in the edge information matching process exceeds a predetermined threshold ( S6, S10).

  For the image regions matched in S2, matching processing other than the edge (for example, matching using the color difference between two target images) is performed (S3). If the matching does not match in S3, estimation is performed based on the machine usage history information and color information, the image is converted, and the matching process is tried again (S5).

  If they match in S3, the position coordinates of the format for one image area itself and the image frame are stored as a common format. In addition, when the image conversion is performed in S4 and S5 and they match, the conversion contents are also stored. That is, common format information, information on the image conversion method to the common format, difference information other than the common format, and position information of each region with respect to the image frame are accumulated (S9).

  Even if rematching is performed from S5 to S3, for an image area where the number of mismatches in the matching process other than the edge information exceeds a predetermined threshold, the image after image conversion having the best result in the matching process in S3 In the area, color information matching processing that is looser than the standard of S3 is performed (S7, S8). If the matching in S8 does not match, the two compared image areas have completely different color information, and the image area itself and the image area for the image frame are used as difference information, as in the case where the matching of edge information does not match. Are accumulated (S10). Also, if they match in the matching in S8, the color information of both is roughly matched, so that the position coordinates of the format for one image region and the image frame are stored as a common format as in S9. However, the difference from S9 is that auxiliary data for assisting color information matching is also accumulated (S11).

  According to the present invention, by saving only common format information (template information) and conversion information, it is not necessary to save separately for templates that have been edited or processed, and the capacity of the storage device is reduced. be able to.

The processing flowchart of the Example of this invention is shown. 1 shows an overall configuration of a digital image processing apparatus (MFP) according to the present invention. The structure of the 1st, 2nd image processing part is shown. The example of an image explaining the process of Example 1 is shown. It is a figure explaining the matching process of Example 1. FIG. An example of hue distribution information of image A is shown. It is a figure explaining the conversion method by a parameter. Indicates information stored in the HDD. The example of an image explaining the process of Example 2 is shown. Shows machine usage history information. It is a figure explaining the matching process of Example 2. FIG. It is a figure explaining the process which cuts out and accumulate | stores a part of image. An example of the screen of the operation display device during the HDD capacity reduction operation is shown. The example of an image explaining the process of Example 3 is shown. Indicates editing history information related to color information. It is a figure explaining the matching process of Example 3. FIG. An example in which similar images are mapped is shown.

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

  An embodiment in which the image processing apparatus of the present invention is applied to an MFP (multifunction machine) that combines a FAX (facsimile) function, a printer function, a scanner function, a document box function and the like based on a digital color copying machine will be described below.

  FIG. 2 shows the overall configuration of a digital image processing apparatus (MFP) according to the present invention. As shown in FIG. 2, the MFP includes a reading device 101, a first image processing unit 102, a bus control device 103, a second image processing unit 104, an HDD 105, a CPU 106, a memory 107, and a plotter. I / F device 108, plotter device 109, operation display device 110, line I / F device 111, external I / F device 112, S.I. B. 113, ROM 114, general-purpose bus 115, and the like. This MFP is configured to be able to perform data communication with the FAX 116 and the PC 117 which are external devices.

  The reading device 101 includes a line sensor composed of a CCD photoelectric conversion element, an A / D converter, and a drive circuit thereof. Based on the density information of the original obtained by scanning the set original, the 8-bit RGB digital data is obtained. Image data is generated and output to the first image processing unit 102.

  The first image processing unit 102 performs processing for unifying the digital image data input from the reading device 101 to a predetermined characteristic and outputs the digital image data. The characteristics to be unified are characteristics suitable for changing the output destination when image data is stored in the MFP and then reused, and details thereof will be described later. In this embodiment, the image processing is executed by a DSP.

  The bus control device 103 is a data bus control device that transmits and receives various data such as image data and control commands necessary in the image processing device, and has a bridge function between a plurality of types of bus standards. In this embodiment, the bus control device 103 is connected to the first image processing unit 102, the second image processing unit 104, and the CPU 106 via a PCI-Express bus, and is connected to the HDD 105 via an ATA bus. It has become.

  The second image processing unit 104 performs image processing suitable for an output destination designated by the user on the image data whose characteristics predetermined by the first image processing unit 102 are unified, and outputs the image data. Details thereof will be described later.

  An HDD (storage device) 105 is a large-sized storage device for storing electronic data that is also used in a desktop personal computer, and mainly stores image data. In this embodiment, an ATA bus-connected hard disk standardized by expanding IDE is used. Since image data is larger in capacity than ordinary text files, storing a large amount of image data will fill up the storage device and not be able to store new image data. Influence.

  The CPU 106 is a microprocessor that controls the entire control of the digital image processing apparatus. In this embodiment, an integrated CPU in which a + α function is added to a CPU core that has been popular in recent years is used. For example, a connection function with a general-purpose standard I / F or a crossbar switch is used in a PMC RM11100. A CPU in which these bus connection functions are integrated is used.

  The memory 107 temporarily stores data in order to absorb the speed difference when bridging between multiple types of bus standards and the processing speed difference of the connected components themselves, and the CPU 106 stores the digital image processing apparatus. It is a volatile memory that temporarily stores programs and intermediate processing data when performing control. Since high speed processing is required, the CPU 106 starts up the system with a boot program stored in the ROM 114 during normal startup, and thereafter performs processing with a program developed in a memory accessible at high speed. In this embodiment, a standardized DIMM used for a personal computer is used as the memory 107.

  When the plotter I / F device 108 receives CMYK image data sent via the general-purpose standard I / F integrated with the CPU 107, the plotter I / F device 108 performs a bus bridge process for outputting to the dedicated I / F of the plotter device 109. . In the present embodiment, for example, a PCI-Express bus is used as the general-purpose standard I / F.

  When the plotter device 109 receives the image data composed of CMYK, the plotter device 109 outputs the received image data on the transfer paper using an electrophotographic process using a laser beam. S. B. Reference numeral 113 denotes one of chip sets used for a personal computer, which is a general-purpose electronic device called South Bridge. S. B. Reference numeral 113 denotes a bus bridge function that is often used when constructing a CPU system mainly including a PCI-Express and an ISA bridge, and bridges the ROM 114 in this embodiment. The ROM 114 is a memory that stores a program (including a boot program) used when the CPU 106 controls the digital image processing apparatus.

  The operation display device 110 is a part that performs an interface between the digital image processing device and the user. The operation display device 110 includes an LCD (liquid crystal display device) and key switches, displays various states and operation methods of the device on the LCD, and receives a user's request. Detects key switch input. In this embodiment, the CPU is connected to the CPU via a PCI-Express bus. The line I / F device 111 is a device that connects a PCI-Express bus and a telephone line. With this apparatus, the digital image processing apparatus can exchange various data via the telephone line. The external I / F device 112 is a device that connects a PCI-Express bus and an external device. This device enables the digital image processing apparatus to exchange various data with the external device. In this embodiment, a network (Ethernet (registered trademark)) is used for the connection I / F. That is, this digital image processing apparatus is connected to the network via the external I / F apparatus 112. A FAX 116 is a normal facsimile, and transmits / receives image data to / from the MFP via a telephone line. A PC 117 is a personal computer, and a user performs various controls and input / output of image data with respect to the digital image processing apparatus via installed application software and drivers.

The basic normal copy operation processing / transmission operation (when capacity reduction is not performed) will be described below, and then the capacity reduction operation will be described.
(Copy operation)
The user sets an original on the reading device 101 and performs setting of a desired mode and input of copy start using the operation display device 110. The operation display device 110 converts information input by the user into control command data inside the device and issues it. The issued control command data is notified to the CPU 106 via the PCI-Express bus.

  The CPU 106 executes a copy operation process program in accordance with the copy start control command data, and sequentially executes settings and operations necessary for the copy operation. The operation process is described below.

  # 1. The RGB 8-bit image data obtained by scanning the document with the reading device 101 is unified with predetermined characteristics by the first image processing unit 102 and sent to the bus control device 103.

  # 2. When the bus control device 103 receives the RGB image data from the first image processing unit 102, the bus control device 103 stores it in the memory 107 via the CPU 106.

  # 3. Next, the RGB image data stored in the memory 107 is sent to the second image processing unit 104 via the CPU 106 and the bus control device 103.

  # 4. The second image processing unit 104 converts the received RGB image data into CMYK image data for plotter output and outputs it.

  # 5. When the bus control apparatus 103 receives CMYK image data from the second image processing unit 104, the bus control apparatus 103 stores the data in the memory 107 via the CPU 106.

  # 6. Next, the CMYK image data stored in the memory 107 is sent to the plotter device 109 via the CPU 106 and the plotter I / F device 108.

  # 7. The plotter device 109 outputs the received CMYK image data to transfer paper, and a copy of the document is generated.

  FIG. 3A shows the configuration of the first image processing unit 102, and FIG. 3B shows the configuration of the second image processing unit 104.

  In the first image processing unit 102 in FIG. 3A, the image area separation processing unit 201 is a character area or a pattern (halftone dot) area for each pixel of the RGB image data received from the reading apparatus 101. And the area information is used in the filter processing described later. The background removal processing unit 202 calculates the background portion of the RGB image data received from the reading apparatus 101 while following each line, and performs processing for removing an appropriate background value. In addition to normal white paper originals, backgrounds such as newspapers and colored papers are removed to produce a white original.

  The γ conversion unit 203 unifies the brightness of the RGB image data received from the reading device 101 into predetermined characteristics. In this embodiment, the lightness is converted into linear characteristics, and gradation correction, gray balance correction, and the like are mainly performed. The isolated point removal processing unit 204 detects and deletes isolated points in the image data.

  The filter processing unit 205 performs a process for improving the sharpness on the character part of the RGB image data in accordance with the image area separation information obtained by the image area separation process. Part) is subjected to a smoothing process and a process emphasizing smooth gradation.

  The color conversion unit 206 unifies the colors of the RGB image data with predetermined characteristics. In this embodiment, conversion is performed so that the color space is a color space such as Adobe (registered trademark) -RGB. The scaling processing unit 207 unifies the size (resolution) of the RGB image data to a predetermined characteristic. In this embodiment, the size (resolution) is converted to 600 dpi.

  In the second image processing unit 104 in FIG. 3B, the filter processing unit 301 corrects the sharpness of the RGB image data so that the reproducibility when output to the plotter device 109 is improved. Specifically, sharpening / smoothing processing is performed according to desired mode information. For example, in the character mode, a sharpening process is performed to make the character clear / clear, and in the photo mode, a smoothing process is performed to smoothly express gradation.

  When the color conversion unit 302 receives 8-bit RGB data, the color conversion unit 302 converts the data into 8-bit CMYK, which is a color space for the plotter device. At this time, the saturation is also adjusted according to the mode information desired by the user. The scaling processing unit 303 performs size (resolution) conversion on the size (resolution) of the CMYK image data in accordance with the reproduction performance of the plotter device 109. In this embodiment, since the performance of the plotter 109 is 600 dpi output, no particular conversion is performed.

  When the gradation processing unit 304 receives 8 bits for each of CMYK, it performs gradation number conversion processing in accordance with the gradation processing capability of the plotter device 109. In this embodiment, the number of gradations is converted into 2 bits of CMYK using an error diffusion method which is one of pseudo halftone processes. In the mask processing unit 305, the peripheral data of the image data is overwritten with the white data in order to erase the shadow of the document from the gradation-converted CMYK data.

  FIG. 1 is a process flowchart of an embodiment of the present invention. Hereinafter, the capacity compression operation, which is a feature of the present invention, will be described with reference to FIG. Note that the image storage operation is that the CPU 106 stores the image data stored in the memory 107 in # 2 in the HDD 105 via the bus control device 103 after the operations from # 1 to # 2 of the copy operation described above. It is an operation.

  In order to reduce the capacity of the HDD 105, a case where the capacity of the images A and B stored in the HDD 105 in the image storing operation is reduced will be described. The trigger for the capacity reduction operation described below is performed when the HDD 105 exceeds a predetermined capacity, for example, in the case of an MFP equipped with the HDD 105 having a capacity of 500 GB, when the used capacity reaches 480 GB. Such a capacity reduction operation is preferably performed when the MFP is not used for a long time, such as at night or during holidays.

  # 10. The images A and B stored in the HDD 105 are expanded in the memory 107 via the bus control device 103.

  # 11. The CPU 106 performs image area division on the images A and B (S1). As a method of dividing the image region, layout analysis described in Patent Documents 2 and 3 is used. The upper left X1, Y1 coordinates of all the image areas divided and the lower right X2, Y2 coordinates of the image areas are stored in the memory 107. 4A shows image data A (full color image), FIG. 4B shows image data B (monochrome image), FIG. 4C shows image data A after area division, and FIG. 4D shows after area division. Image data B is shown.

  # 12. Further, binary edge images (contour images) of the image data A and the image data B are calculated. This processing may be performed by the CPU 106, or the function of the filter processing unit 301 of the second image processing unit 104 may be used from the memory 107 via the bus control device 103.

  The CPU 106 performs pattern matching processing of edge images between the respective image areas by brute force (S2). Since the image data A and B are both scanner images, even if the original templates are the same, they cannot completely match. Therefore, if the size of the image region is within ± 5%, for example, the scaling process is performed to make the two image regions the same size, and then matching with edge information is performed (the numerical value of 5% is arbitrary) Yes, the user may choose as appropriate).

  In the matching process, if the number of matching mismatched pixels is within a predetermined threshold with respect to the total number of pixels in the image area, the two image areas are assumed to match. For example, as a result of matching between binary edge image areas of 10000 pixels, 300 pixels of them are mismatched (the pixels are white and black), and if the predetermined threshold is 5%, 10000 × Since 5% = 500 pixels or less, it is determined that the image regions match each other. Only the image areas matched by edge information matching are processed after # 13. FIGS. 4E and 4F show the image areas of the image data A and B that are matched by the edge information. That is, the picture 11 of the image data A and the picture 21 of the image data B match, the picture 12 of the image data A and the picture 22 of the image data B match, and the picture 13 of the image data A and the picture 23 of the image data B match. Match.

  # 13. In step # 12 described above, the CPU 106 performs color information matching processing on the image area where the edge information matches (S3). For example, if the image data A and B are RGB data defined in the Adobe (registered trademark) -RGB color space, the RGB space is converted into the Lab space and the color differences are compared. The color difference is the Euclidean distance between two points in the Lab space, and there is a certain degree of correlation between the difference in human appearance and the distance. Alternatively, the Euclidean distance may be evaluated by ΔE94 color difference that is corrected by applying a predetermined weight.

  In two image areas to be compared for color difference, the color difference for each pixel is obtained, and if the average color difference / maximum color difference in the area is within a predetermined threshold value, the color information of the two image areas match. judge. For image areas in which the average color difference and the maximum color difference are matched within a predetermined threshold in the color information matching process, the process proceeds to # 18 described later.

  However, since the image A is actually a full-color image and the image B is a monochrome image, even if color information is simply compared, they do not match. If the color information does not match even if the color information is compared, the edge information is the same and the original color information was the same, but it was determined that the color information no longer matched after performing some image processing or editing processing, and that processing Inferring the contents, estimating the method of converting from image A to image B (or the method of converting from image B to image A), and matching or mismatching in the matching process by the estimation process, image conversion, and color information matching process Is repeated until the number of times reaches a predetermined threshold (S3, S5).

  FIG. 5 is a diagram for explaining the matching process when the color information does not match.

# 14. An estimation process for estimating the processing contents when the color information does not match will be described.
"Color information estimation process" (first half of S5)
First, based on the color difference comparison information result, the CPU 106 checks the processing content for the image.

Comparison of color difference for each hue It is checked whether a degree of color difference mismatch occurs for each hue. First, paying attention to the image A having a wide range existing in the Lab space, each pixel in the image area of the image A is divided into, for example, six hues based on the a and b values, and any of the twelve hues. Check for discrepancies.

  In the case of the image A and the image B, since the difference is between a full color image and a monochrome image, the color difference is different in all hues. If the hues of all hues are different, check the hue distribution of both images. Then, as shown in FIGS. 6A and 6B, in the image A, there are images having all hues and a certain degree of saturation, but in the image B, the values of a and b are 0 for all the pixels. Therefore, it can be estimated that the image A is a full-color image and the image B is a monochrome image.

  The number of pixels for each hue shown in FIG. 6 is an ideal value, but since it is actually an image read by a scanner, there are pixels that are not counted as low-saturation pixels even in a monochrome image, Even if the same document is read, the same number of pixels distribution is not obtained. Therefore, a predetermined threshold value is set, and if it is less than the threshold value, it is determined that the image is a monochrome image. Show in case).

In addition, when the colors of all hues are different, various estimations can be made based on the distribution of hues of two images. An example is shown below.
It can be estimated that negative-positive (black and white) conversion has been performed on images having complementary colors as shown in FIGS.
As shown in FIGS. 6A and 6D, pixels are reduced from each hue by the same ratio with respect to the image A (the order of increasing hues does not change), so saturation conversion (light / dark) ), Toner is saved when the document is output, etc. (a method for narrowing down to any one of a plurality of candidates will be described in the third embodiment).
When the hue distribution is moving as shown in FIGS. 6A and 6E (R in (a) = Y in (e), Y in (a) = G in (e)) It can be presumed that editing is in progress.

Further, it is possible to estimate the case where the color difference is different only in a specific hue instead of the entire hue, and an example is shown below.
As shown in FIGS. 6 (a) and 6 (f), if image A has pixels of all hues but one image has only R hues, it can be estimated that two-color conversion has been performed. it can.
If the image A has pixels of all hues but one image does not have a specific hue, it can be estimated that a drop color (achromatic) process is being performed.

  As described above, two image conversion methods are estimated based on color information.

  # 15. The image is converted using the conversion method estimated in # 14 (second half process of S5).

  In the case of the image A and the image B, as described above, since it can be estimated that the monochrome conversion has been performed from the image A to the image B, the image A is actually subjected to the monochrome conversion. The monochrome conversion is performed by the CPU 106 or the second image processing unit 104. An image obtained by monochrome-converting the image A is referred to as an image A ′.

  # 16. Color information matching processing is performed on the image converted in # 15. That is, the image areas matched with the edge information in the images A ′ and B are again subjected to color information matching processing to confirm whether the average color difference and the maximum color difference are within a predetermined threshold. If the matching process does not match, the estimation process of # 14 and the conversion process of # 15 are further repeated, and the matching process is performed again. Then, in an image area where the number of mismatches exceeds a predetermined threshold, the following process of # 17 is performed. In an image area where the number of mismatches is less than the predetermined threshold and matching of color information matches, the process of # 18 is performed. (S7).

  # 17. Processing when color information matching does not match will be described. First, it is determined whether or not the average color difference is within a second threshold value that is larger (loose) than the matching criterion (S8). For example, when the image regions “Picture 11” and “Picture 21” in FIGS. 4C and 4D are compared with each other, the best result among the image conversions performed several times is an average color difference of 4.5, Assume that the maximum color difference is 7, and the reference threshold value that is matched in the color information matching is 3 for the average color difference and 5 for the maximum color difference. An image region obtained by converting the “picture 11” at this time is referred to as “picture 11 ′”. In this case, in the matching between the image areas, the average color difference and the maximum color difference are both equal to or larger than the threshold value, and thus the matching is not matched. When the number of mismatches exceeds a predetermined threshold value, the average color difference and the maximum color difference value with the best results among the image conversions are compared with the second threshold value. Assuming that the second threshold is an average color difference of 5 and a maximum color difference of 8, the matching between the image areas “Picture 11” and “Picture 21” is within the second threshold.

  When the second threshold value is exceeded, the color information is completely different in the two image areas, and conversion between the two image areas is difficult, so the color information of the two image areas is stored separately in the HDD 105. To do. On the other hand, if it is within the second threshold, it is difficult to convert some pixels, but it is determined that conversion of many pixels is possible, and information for assisting conversion of only some of the pixels is combined. Save in the HDD 105.

Two specific examples of information to assist are given below.
(1) First, since the location exceeding the first maximum color difference threshold value 5 is an unacceptable location, the maximum color difference of “Picture 11 ′” and “Picture 21” obtained by performing the image conversion of “Picture 11” is 5 or more. Extract pixels. Then, it is checked whether or not the average color difference is 3 or less of the reference only with the pixels in the area having the maximum color difference of 5 or less.

  As a result, if the average color difference is 3 or less, only the extracted pixel data having the maximum color difference of 5 or more is stored as a separate plane as it is, and “Picture 11 ′” and “Picture 21” are handled as matched. If it exceeds 3, it is confirmed again whether the average color difference is 3 or less, except for those having a color difference of 4.5 or more instead of 5 or more. If the average color difference still exceeds 3, the color difference is 4 or more. If it still exceeds, the color difference value is gradually reduced to 3.5 or more, and the above process is repeated until the average color difference becomes 3 or less. However, if the second threshold value is set to a low value to some extent, if the maximum color difference threshold value (= 5) is removed, it will be within the average color difference threshold value. This is because the number of pixels exceeding the maximum color difference threshold (= 5) exceeds the maximum color difference threshold (= 5) as compared to storing the color information of the entire image area in another plane because of several pixels at most. This is because the capacity of the HDD 107 is smaller when only the pixels are stored in a separate plane, and when only the conversion method is stored otherwise.

For example, if the same document is output by two printers with different toners, and the two sheets are matched with color information, the gamuts that can be reproduced by the printers are different, so only the colors outside the gamut greatly exceed the maximum color difference. As a result, the average color difference may become a large value. In such a case, only the color gamut outside the gamut is stored in a separate plane, and otherwise the matching is the same, that is, it is stored as a common format, so that it is possible to further reduce the HDD capacity compression. .
(2) Conversion Method Using Parameters Similarly to (1), a pixel having a high average color difference (mainly a pixel exceeding the maximum color difference threshold (= 5)) is extracted. For the extracted pixel, a parameter for converting from “Picture 11 ′” to “Picture 21” is obtained for each hue by a primary conversion formula (for example, 12 hue division). Each pixel extracted from “Picture 21” is set as a target value, and an optimal parameter is obtained by regression calculation (least-squares method) or the like as an input value for each primary transformation of “Picture 11 ′” corresponding thereto. FIG. 7 is a diagram for explaining a conversion method using parameters.

  Using the parameter (coefficient), the color difference between the result of the primary conversion from the pixel value of “Picture 11 ′” and the pixel value of “Picture 21” is obtained, and the average color difference is smaller than the threshold 3 and is the maximum If the color difference is smaller than the threshold value 5, the parameter is stored as auxiliary data. Even if the average color difference is smaller than the threshold value, if it exceeds the threshold value for the maximum color difference, the threshold values for both the average color difference and the maximum color difference can be obtained by performing optimization with constraints on the maximum color difference. The parameter is calculated so as to clear Optimization with constraints on the maximum color difference usually involves calculating a parameter that minimizes the evaluation function that minimizes the color difference, but this evaluation function imposes a penalty when the maximum color difference exceeds a certain threshold. By optimizing the evaluation function to be small by iteration, it is possible to perform optimization with a restriction on the maximum color difference. Even if this optimization is performed, if either the average color difference or the maximum color difference does not satisfy the threshold value, only the hue may be subjected to secondary conversion.

  In addition, since this primary conversion is executed as a hue division linear masking process during the color conversion processing of the first and second image processing units, when actually obtaining the picture 21 from the parameters of the primary conversion, Pixels for which the maximum color difference of the picture 11 ′ exceeds the threshold value 5 are transmitted from the memory 107 to the second image processing unit 104 via the bus control device 103. The stored parameters are also set in the masking parameter register of the color conversion process of the second image processing unit 104. By using the second image processing unit 104 in this way, the load of the CPU 106 is distributed and the processing speed is increased. If the second image processing unit 104 is an ASIC or the like, the processing is overwhelmingly faster than the processing by the CPU 106.

  # 18. As for the image areas matched in the color information matching process, only one image area with more information is stored in the HDD 107 as a common format. The position coordinates of each image area are also stored together (S9, S11). Further, when the image conversion estimation method is used for conversion and matching is performed, the conversion method is also stored in the HDD 107 together. In step # 17, if some auxiliary data is necessary to perform color information matching processing, the auxiliary data is also stored together with the common format. For image areas that do not match in the color information matching process, the image area itself and the position information are stored together.

  FIG. 8 shows information stored in the HDD in the case of images A and B. Rather than storing image data A and image data B, the storage of common format information, difference information, and conversion information eliminates redundant data as shown in FIG. . In order to create the image data B from the data stored in the HDD 107 in FIG. 8, the character 21 and the character 22 are first synthesized with the position coordinates stored in the monochrome format of the common format information. Becomes image data B.

  In this embodiment, the image conversion method is added to the common format data and its difference data, and the conversion method to the common format (and auxiliary data for the converted data) is stored together, thereby reducing the capacity compared to the conventional method. Can be reduced. Also, if color information conversion is not possible and there is a mismatched part in the second matching process, the information for correcting only the mismatched part is saved together, so that the color information of the template information is saved in a separate plane. Also, the capacity can be reduced. Further, by making the correction information difference data from the conversion target data, the correction information can be easily obtained and can always fall within a predetermined allowable range of mismatch. Further, by making the correction information a linear expression or a quadratic conversion expression, the capacity can be made smaller than when the image data is stored when the number of pixels in the mismatched area is large. In addition, by using color information, it is possible to estimate an image conversion method related to color. For example, two-color conversion, monochrome conversion, negative / positive conversion, designated color erasure conversion, CUD (color universal design) compatible conversion, toner save conversion, and the like can be given. Furthermore, it is possible to efficiently estimate an image conversion method by performing color information matching on edge information matching, that is, having the same shape.

  FIG. 9 is a diagram for explaining the processing of the second embodiment. In order to reduce the capacity of the HDD 105, processing for reducing the capacity of the image C (FIG. 9A) and the image D (FIG. 9B) stored in the HDD 105 in the image storage operation will be described. In the following, processing that is different from the first embodiment will be mainly described.

  # 20. # 21. This is the same as the processing of # 10 and # 11 of the first embodiment.

  # 20. # 21. Similar to the first embodiment, matching processing is performed using edge information (S2). In the case of image C and image D, all the image areas do not match, so the CPU 106 calls the machine editing information and usage history information in the memory 107 (if even one image area matches, step # 23) Go to).

  FIG. 10 shows machine usage history information. Of the editing information in FIG. 10, it is estimated which image C or image D is being edited in order from the editing information having the highest probability of being used. Referring to FIG. 10, it is estimated that the two processes are performed because the 2-in-1 consolidated printing and the staple usage rate are high. Whether or not the image C or the image D is subjected to the aggregate printing is estimated from information such as the arrangement position (position information) of the image area, the blank of the boundary portion of the aggregate image, and the size of the characters (for example, Patent Documents). 4). Then, it can be seen that there is a high possibility that the image D is performing 2-in-1 consolidated printing.

  Whether or not stapling is performed may be determined by searching whether or not there are staple marks only in the left and right and upper and lower peripheral portions of the image. The same applies to the punch hole search. Even if the staple mark search is performed, the image C and the image D do not show the mark, so the image conversion relating to the staple mark is not performed.

  # 23. For this reason, as shown in FIG. 11, the image C is halved and the matching process using the edge information is performed again. In addition, the size of the image C is as it is, and there is no problem even if the image D is doubled and the matching process is performed (which process is performed will be described later).

  In this way, estimation is performed using the positional relationship between the editing information and the image area as a hint, image conversion is performed, and edge information matching processing is repeated. An image area in which the number of mismatches in this matching process exceeds a predetermined threshold value stores the image area itself and its position coordinates as difference information (S6). If one image region does not match in the matching process, combining the image C and the image D is not effective in reducing the capacity compression, and therefore other images are combined. In the matching processing of FIGS. 9C and 9D, the result is picture 31 = picture 41 = picture 44, picture 32 = picture 42 = picture 45, picture 33 = picture 43 = picture 46.

  The reason for adopting the above processing is that the input manuscript is originally output by the same MFP, but even if it is output by another MFP, the usage history of the editing function is, for example, in the same department or in-house. This is because there are many similar cases.

  In the present embodiment, the edit history at the time of output is used as the edit history, but naturally the edit setting at the time of image accumulation may be used. When storing images, it is basic to store redundant data without editing, but it may be desirable to delete images when trimming, etc. In that case, using information such as the trimming position, Perform the matching process.

  For example, as shown in FIG. 12, when the image of the dotted line part extracted from the image A and the image A itself are matched with edge information, the image areas of the characters 12 and 13 are matched by the edge information. Although they match, a part of the picture 12 is cut (referred to as picture 12 ′), and therefore does not match in the matching process using the edge information. Therefore, when the image accumulation of the dotted line portion is performed, the position coordinates of the part to be cut out and the image size before the cutting are stored. When matching with the image A during the capacity compression reduction operation, the image A is cut out using the cut position coordinates, and the matching process with the picture 12 ′ can be matched.

  # 24. The subsequent processing is the same as in the first embodiment. However, in the second embodiment, the common format information may be “picture 31, picture 32, picture 33” and “conversion method to image D: 1/2 scaling” (referred to as common format A). “Picture 41 (Picture 44), Picture 42 (Picture 45), Picture 43 (Picture 46)” and “Conversion method to image C: Double scaling” (referred to as common format B) may be used.

  Basically, color information matching processing is performed with the image D doubled. Therefore, if there is a margin in the threshold values of the average color difference / maximum color difference, which are the criteria for matching processing, color information matching processing is performed between the image C halved and the image D. In this case, as long as the average color difference / maximum color difference threshold values are satisfied, the common format may be any of A and B described above. However, since A has a larger size (or resolution) as a common format, it contains more image information than B, resulting in higher image quality. On the other hand, when B is selected, the capacity can be greatly reduced accordingly.

  Further, even in the color information matching process in the case of the common format B (the image C is ½ and the image D is as it is), the matching process is performed when there is a clear margin than the average color difference threshold value and the maximum color difference threshold value. The resolution of the image area of the image C is reduced (for example, ¼) until the result reaches the threshold values of the average color difference and the maximum color difference. Then, by accumulating the image area of the image C with the reduced resolution (data of 1/4 resolution) as a common format, the capacity of the common format can be reduced as much as possible.

  FIG. 13 shows an example of the screen of the operation display device during the HDD capacity reduction operation, and the administrator can select which one of priority should be given to high image quality or capacity reduction by the operation display device 110. Further, if low image quality is selected, the average color difference / maximum color difference threshold may be increased.

  In the present embodiment, convenience can be improved by selecting a balance (tradeoff) between the capacity and the conversion accuracy (matching rate) of the image data. In addition, the balance (tradeoff) between the capacity and the conversion accuracy (matching rate) of the image data can be selected in multiple stages, so that convenience is improved. By using the positional information and size of the image area, it is possible to estimate the image conversion method related to the editing function. For example, an aggregation function (2 in 1, 4 in 1, etc.), a scaling function, a trimming function, a punch hole / staple mark, a mirror function, and the like can be given. Furthermore, when image data is input, it is possible to estimate an image conversion method efficiently when the image data is edited. Basically, as a system for estimating the image conversion method, it is best to store the image data in the image storage means as it is without performing image processing as much as possible (that is, to store the image data). For example, when only a part of the image data is necessary, it is a waste of capacity even if the other extra part is saved, which is not good from the viewpoint of security. Therefore, it is better to store the image data after editing.

  FIG. 14 is a diagram for explaining the processing of the third embodiment. In order to reduce the capacity of the HDD 105, processing for reducing the capacity of the image E (FIG. 14A) and the image F (FIG. 14B) stored in the HDD 105 in the image storage operation will be described. The image F has a lighter color than the image E (low saturation and high brightness). Since the processing other than the estimation processing of # 14 is the same as that of the first embodiment, only the estimation processing of the third embodiment will be described below.

  Even if only the hue distribution is compared as in the first embodiment, the pixels of the image data E and F are distributed in all hues, and the conversion process cannot be estimated only by the hue distribution. Therefore, the CPU 106 expands editing history information related to color information in the memory 107.

  FIG. 15 shows editing history information related to color information. In the edit history information shown in FIG. From FIG. 15, since the usage rate of the toner save mode is the highest, the distribution of the images E and F in the Lab space is viewed in detail from the viewpoint that toner save is performed on one image.

  In other words, in a document that has been subjected to toner save, the amount of toner on the pattern portion or the like is less than usual, so the brightness is low if the paper is low in saturation and white. Comparing the entire pixels of the image E and the image F, the image F tends to be lower in saturation and lighter than E. Therefore, it is estimated that the image F is an image that has been subjected to the toner save mode, and the image E is a toner. It converts into the RGB image of a save mode (refer FIG. 16).

  Specifically, the image E developed in the memory 107 is transferred to the second image processing unit 104 via the bus control device 103, where normal plotter output processing (including toner save processing) is performed. The image is converted into a CMYK image and stored in the memory 107 again. The CPU 106 obtains RGB values when a CMYK image output on white paper is read by a scanner (using a scanner automatic calibration function or the like), and performs color information matching processing between the RGB values and the image F. (See FIG. 16).

  If there is a mismatch in this matching process, the process with the next highest usage rate in the editing history is listed as a candidate. If the comparison result of the pixels in the Lab space and the tendency of the pixels in the editing process match, the image is converted to the image and the color information matching process is performed. This processing loop is repeated until the image areas match or the number of mismatches reaches a predetermined threshold.

  CUD (Color Universal Design) support, which has the next highest editing history after toner save, is a conversion to a color with good visibility when the combination of adjacent color combinations is poor. Therefore, in both the image E and the image F, the color difference between adjacent colors in each pixel is obtained, and the magnitude of the color difference between the images E and F is compared. If there are many places where the color difference is extremely large, it is estimated that the image is CUD-compliant, an image that is not CUD-compliant is converted to CUD-compliant, and color information matching processing is performed.

  In this embodiment, by using the user's editing history, it is possible to perform estimation more efficiently than estimating the image conversion method using only color information and position information.

  In the first to third embodiments, the two images that have already been selected are converted into the common format and the difference information. In the present embodiment, the method for selecting the two images is described. .

  As described above, the operations of the first to third embodiments are performed when a large amount of image data is accumulated in the HDD 105 and the capacity becomes scarce. For this reason, by selecting two images that are likely to have a common format from among a large number of image data stored in the HDD 105, the processing time of the HDD compression reduction operation is shortened. The fourth embodiment will be described below by dividing the processing during image storage and the HDD compression reduction operation.

(At the time of image accumulation operation)
As described above, copy operations # 1 and # 2 are performed. For the image data stored in the memory 107 in # 2, the CPU 106 stores some (at least one) of the following information in the HDD 105 together.
-Creator information User information may be input by the operation display device 110, or a reader or the like that reads biometric information may be attached to the expansion bus, and an individual may be specified by biometric authentication. Further, not only the name of the user but also the group (team, section, department, etc.) to which the user belongs may be stored together.
Character information in the image Using the layout analysis and OCR processing described above for the image data stored in the memory 107, characters such as titles and footers / headers are extracted and stored as keywords. To do.
Image Type Information The operation display device 110 stores information that can specify to some extent the type of document to be accumulated, such as a photo document mode, a map mode, and a highlighter mode. Even if it is not information from the operation display device 110, images such as layout information, spatial frequency, etc. read from the scanner, such as how to use colors (whether there are many chromatic colors), whether there are many pictures, or many characters. Use the feature quantity. For example, similar images are mapped as shown in FIG.

(During pressure reduction operation)
When the compression reduction operation trigger is transmitted to the CPU 106, such as when the HDD capacity (usage capacity) exceeds 95%, the CPU 106 expands the information accumulated in the image accumulation operation as described above in the memory 107, and uses that information as a basis. Select two images.
-Creator information Since the same creator is likely to use the same format, the compression reduction operation is performed between images having the same creator. In addition, it is highly likely that not only the same creator but also the same team and the same section will use the same format, so priority is given to it.
-Character information in images If titles, footers, headers, and the like are the same, there is a high possibility that they are in the same format, so the compression reduction operation is performed between the same images.
Image Type Information Since the more similar image types are, the more likely they are in the same format, the compression reduction operation is performed between the same images. That is, if the format is the same, it is highly possible that the contents of the document are similar. For example, as shown in FIG. 17, if the mapping is performed, the compression reducing operation is performed between the coordinates close to each other. If the CPU 106 can refer to a plurality of pieces of the above-described three types of information, the CPU 106 increases the priority order for the images having as many common points as possible, and performs the compression reducing operation.

  When the number of image data stored in the image storage means is large, it takes time to perform the matching process for all image data in a round-robin manner. In this embodiment, the processing time can be shortened by selecting an image to be subjected to matching processing. In addition, if the character information of the image data such as title, header, and footer is the same or similar, it is likely that the same template or format is used, so efficient matching processing can be performed. It becomes possible. Further, if the feature amount of the image data is close, there is a high possibility that the same template or format is used, so that efficient matching processing can be performed. Furthermore, documents created by the same creator or files created by people belonging to the same group are more likely to use the same template or format, which is efficient. Matching processing can be performed.

  According to the present invention, a storage medium in which a program code of software for realizing the functions of the above-described embodiments is recorded is supplied to a system or apparatus, and a computer (CPU or MPU) of the system or apparatus is stored in the storage medium. This is also achieved by reading and executing the code. In this case, the program code itself read from the storage medium realizes the functions of the above-described embodiment. As a storage medium for supplying the program code, for example, a hard disk, an optical disk, a magneto-optical disk, a nonvolatile memory card, a ROM, or the like can be used. Further, by executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also an OS (operating system) operating on the computer based on an instruction of the program code. A case where part or all of the actual processing is performed and the functions of the above-described embodiments are realized by the processing is also included. Further, after the program code read from the storage medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, the function expansion is performed based on the instruction of the program code. A case where the CPU or the like provided in the board or the function expansion unit performs part or all of the actual processing and the functions of the above-described embodiments are realized by the processing is included. Further, the program for realizing the functions and the like of the embodiments of the present invention may be provided from a server by communication via a network.

DESCRIPTION OF SYMBOLS 101 Reading apparatus 102 1st image processing part 103 Bus control apparatus 104 2nd image processing part 105 HDD
106 CPU
107 Memory 108 Plotter I / F Device 109 Plotter Device 110 Operation Display Device 111 Line I / F Device 112 External I / F Device 113 B.
114 ROM
115 General-purpose bus 116 FAX
117 PC

JP 2008-146104 A JP 2008-146104 A Japanese Patent No. 3278471 JP 2009-71630 A

Claims (18)

  Image input means for inputting an image, image storage means for storing the image, area dividing means for dividing the image into a plurality of areas, and a method for converting an image between a plurality of images stored in the image storage means In the image processing apparatus comprising: an estimating unit that estimates the image area; and a matching processing unit that performs pattern matching processing on the image areas divided by the area dividing unit, the estimating unit includes: The image conversion method between the two images is estimated, the first image is converted by the estimated conversion method, and the matching processing means is configured to convert the first image converted by the estimated conversion method. The image area of the second image and the image area of the second image are matched, and the common image area information matched by the matching process and the difference other than the common image area information The image processing apparatus characterized by storing information and transform information the estimated in the image storage means.   The matching processing means includes a first matching process for matching the first elements of the first and second images, and an image region matched by the first matching process with respect to the first, Performing a second matching process for matching the second elements of the second image, and creating a correction information for correcting only the mismatched part when there is a mismatch in the second matching process; The common image region information matched in the first matching process, difference information other than the common image region information, the estimated conversion information, and the correction information are stored in the image storage unit. Item 6. The image processing apparatus according to Item 1.   In the case of mismatch in the second matching process, the difference information of the mismatched portion between the first image converted by the estimated conversion method and the second image is used as the correction information. The image processing apparatus according to claim 2.   A parameter of a conversion equation for converting the first image converted by the estimated conversion method into the second image when there is a mismatch in the second matching processing is used as the correction information. The image processing apparatus according to claim 2.   According to the result of the matching process performed on the first image and the second image, the matching processing unit uses either the first image or the second image as common image area information stored in the storage unit. The image processing apparatus according to claim 1, wherein the image processing apparatus is selected.   In the processing result by the matching processing unit, when there is a margin beyond the matching allowable range, the common image area information stored in the image storage unit is deteriorated to the extent that it falls within the allowable range. 3. The image processing apparatus according to claim 1, wherein the capacity of the storage means is further reduced.   The image processing apparatus according to claim 1, wherein the estimation unit estimates an image conversion method using color information.   The image processing apparatus according to claim 1, wherein the estimation unit estimates an image conversion method using a position and / or size of an image region.   An editing history information storage unit for storing user editing history information, wherein the estimating unit estimates an image conversion method using the stored user editing history information. The image processing apparatus according to 1.   The image processing apparatus according to claim 1, wherein the estimation unit estimates an image conversion method using editing information when an image is input.   The image processing apparatus according to claim 2, wherein the first matching processing is edge information matching processing, and the second matching processing is color information matching processing.   The image processing apparatus according to claim 1, wherein the matching processing unit includes an image selection unit that selects an image to be a matching process target from the images stored in the image storage unit.   13. The image processing apparatus according to claim 12, wherein the image selection means selects using character information of a part of the image.   The image processing apparatus according to claim 12, wherein the image selection unit performs selection using document type information of an image.   The image processing apparatus according to claim 12, wherein the image selection unit performs selection using image creator information.   An image input step for inputting an image, an image storage step for storing the image, a region dividing step for dividing the image into a plurality of regions, and a method for converting an image between a plurality of images stored in the image storage step In the image processing method comprising: an estimating step for estimating the image region; and a matching processing step for performing pattern matching processing on the image regions divided by the region dividing step, the estimating step includes: An image conversion method between the two images, the first image is converted by the estimated conversion method, and the matching processing step includes the first image converted by the estimated conversion method. The image area of the second image and the image area of the second image are matched, and the common image area information matched by the matching process and the difference other than the common image area information Image processing method characterized by storing information and transform information the estimated in the image storage step.   A program for causing a computer to realize the image processing method according to claim 16.   A computer-readable recording medium having recorded thereon a program for causing a computer to implement the image processing method according to claim 16.

Images