JP5299197B2 - Image processing apparatus, image processing method, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To achieve subtractive color at a high compression rate and with high image quality by enhancing the accuracy of calculating a color number and improving the accuracy of determining a compression format. <P>SOLUTION: The image processor divides image data into a plurality of areas, calculates a color number in each area, selects a representative color value of an area on the basis of the calculated color number or an inputted color number, generates index image data obtained by replacing each color value of the area where the representative color value is calculated with the representative color value, performs image quality evaluation between image data of the area where the index image data are generated and the index image data of the area, selects the calculated color number when an evaluation result corresponds to a preset evaluation value, increases the color number when the evaluation result does not correspond to the preset evaluation result, and selects a representative color value again on the basis of the increased color number. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to an image processing apparatus, an image processing method, and a program.

  In recent years, when compressing image data of an image including characters, symbols, line drawings, photographs, etc., there is a technique for calculating the number of colors of the image data and compressing the image data in a compression format based on the number of colors. The number of colors is calculated using conditions such as the frequency threshold of the histogram of the image data to be compressed, the cluster minimum size of the color set, and the like.

  For example, a basic region including a pixel block that is determined as an individual character from at least a part of image data is defined, and the pixel value of the pixel is represented for each pixel constituting the pixel block included in the defined basic region. Disclosed is a technique for limiting the number of colors to a representative pixel value candidate included in a set of pixel value candidates and determining the number of colors of the pixel block based on the appearance frequency (histogram) of representative pixel value candidates in the pixel block after the limited colorization. (See Patent Document 1).

  A character region including a pixel block that is determined to be a character from at least a part of the image data is defined, and each character region is sequentially selected as a target character region, and values of significant pixels in the target character region in the original image data are determined. Generates a histogram (occurrence frequency), identifies pixel values that appear at a frequency exceeding a predetermined threshold in the histogram, counts the number of the identified pixel values (the number of pixel values in the character portion), and determines the character color A technique for determining the number is disclosed (see Patent Document 2).

  A binarization process is executed by analyzing a histogram distribution indicating the features of the image data, and separated into foreground (character) data and background data. A technique for determining the number of colors of foreground (character) data by analyzing a histogram distribution state, clustering color sets, or the like has been disclosed (see Patent Document 3).

JP 2005260452 A JP 2006-92050 A JP 2006-333175 A

  However, in the techniques such as cited documents 1 to 3, since the number of colors is determined using a standard that is not directly related to the image quality, colors with low occurrence frequency are lost and the number of colors is erroneously determined. There is a problem. Therefore, when setting the compression format according to the number of colors, there is a problem that an inappropriate compression format is set or the compression rate is lowered, and the image quality of the image data after the compression process is lowered. In addition, if the color of an image part of a size that can be perceived by the human eye is a color with a low occurrence frequency, the color of this image part belongs to the representative color and is replaced with another color, or the character image part May be set as the background image.

  In view of the above problems, an object of the present invention is to improve the calculation accuracy of the number of colors, improve the determination accuracy of the compression format, and realize color reduction with a high compression rate and high image quality.

The invention according to claim 1 is an area forming unit that divides image data into a plurality of areas based on a predetermined condition, a color number calculation unit that calculates a first number of colors for each of the areas, and the number of colors based on the first number of colors calculated by the calculation unit, representatives color value selecting unit, the representative color value of each color value of the representative color value has been selected region to elect a representative color value for each of the regions An index image generation unit that generates index image data replaced with the image quality evaluation of the image data of the area where the index image data is generated and the index image data of the area, and the evaluation result is set in advance and image quality evaluation color number selection portion that selects the number of colors is calculated when corresponding to the result, the region in which the image quality evaluation result of evaluation by the color number selection unit does not correspond to a preset evaluation results, the first given the number of colors And a color number increasing section for calculating a second number of colors obtained by adding, the representative color value selecting unit, the region not corresponding to the evaluation result of the evaluation result is set in advance, by the number of colors increases unit The image processing apparatus selects again the representative color value of the area based on the calculated second color number .

  According to a second aspect of the present invention, in the image processing apparatus according to the first aspect, a compression method determination unit that determines a compression method of the area based on the number of colors selected by the image quality evaluation color number selection unit, Is provided.

  According to a third aspect of the present invention, in the image processing apparatus according to the second aspect, when the number of colors in the region is equal to or larger than a preset number of colors, the compression method determination unit performs a plurality of different compressions. A compression method addition determination unit that calculates the data size after compression of the region for each method and determines the compression method of the region based on the comparison result of the data size for each compression method.

According to a fourth aspect of the present invention, in the image processing apparatus according to any one of the first to third aspects, the image quality evaluation color number selection unit includes the image data of the region where the index image data is generated and the image data Either a mean square error with the index image data of the area or a ratio between the image signal and noise is calculated, and the image quality is evaluated based on the calculated value.

  According to a fifth aspect of the present invention, in the image processing device according to any one of the first to third aspects, the image quality evaluation color number selection unit includes the image data of the area where the index image data is generated and the image data An error image with the index image data of the area is calculated, and the image quality is evaluated based on the error image and human visual system characteristics.

  According to a sixth aspect of the present invention, in the image processing device according to any one of the first to third aspects, the image quality evaluation color number selection unit includes the image data of the region where the index image data is generated and the image data The image quality is evaluated based on the sharpness of the edge with the index image data of the area.

The invention according to claim 7 is a region forming step of dividing image data into a plurality of regions based on a predetermined condition, a color number calculating step of calculating a first color number for each of the regions, and the number of colors Based on the first number of colors calculated in the calculation step, a representative color value selection step of selecting a representative color value for each region, and each color value of the region from which the representative color value is selected is represented by the representative color value. Index image generation step of generating index image data replaced with the image quality evaluation of the image data of the area where the index image data is generated and the index image data of the area, and the evaluation result is set in advance The image quality evaluation color number selection step for selecting the number of colors calculated when corresponding to the result, and the region where the evaluation result by the image quality evaluation color number selection step does not correspond to the preset evaluation result A color number increasing step of calculating a second color number obtained by adding a predetermined number to the number of colors of the color, and the region calculated by the color number increasing step for an area where the evaluation result does not correspond to a preset evaluation result. And a representative color value reselecting step of selecting again the representative color value of the area based on the number of colors of 2.

According to an eighth aspect of the present invention, in the image processing method according to the seventh aspect, the compression method determination step of determining the compression method of the area based on the number of colors selected in the image quality evaluation color number selection step. Including.
According to a ninth aspect of the present invention, in the image processing method according to the eighth aspect, the compression method determining step includes a plurality of different compressions when the number of colors in the region is equal to or larger than a preset number of colors. It includes a compression method addition determination step of calculating the data size after compression of the region for each method and determining the compression method of the region based on the comparison result of the data size for each compression method.

According to a tenth aspect of the present invention, there is provided a region forming unit that divides image data into a plurality of regions based on a predetermined condition, a color number calculating unit that calculates a first color number for each region, and the color Representative color value selection means for selecting a representative color value for each of the areas based on the first number of colors calculated by a number calculation means, and each color value of the area from which the representative color value has been selected as the representative color value Index image generation means for generating the index image data replaced with the image quality evaluation of the image data of the area where the index image data is generated and the index image data of the area, and the evaluation result is set in advance An image quality evaluation color number selection means for selecting the number of colors calculated when the image quality corresponds to the above, and an area where the evaluation result by the image quality evaluation color number selection means does not correspond to a preset evaluation result A program for functioning as a color number increasing means for calculating a second color number obtained by adding a predetermined number to the first color number, wherein the representative color value selecting means has the evaluation result set in advance. For the area not corresponding to the evaluation result, the representative color value of the area is selected again based on the second color number calculated by the color number increasing means.
According to an eleventh aspect of the present invention, in the program according to the tenth aspect, the computer determines the compression method of the area based on the number of colors selected by the image quality evaluation color number selection means. It functions as a means.
According to a twelfth aspect of the present invention, in the program according to the eleventh aspect, when the number of colors in the region is equal to or greater than a preset number of colors, the compression method determination unit is configured to store a plurality of different compression methods. And a compression method addition determining unit that calculates the compressed data size of the region and determines the compression method of the region based on the comparison result of the data size for each compression method.

According to the first, seventh , and tenth aspects of the invention, since the number of colors can be selected based on the image quality evaluation, it is possible to prevent under- and over-identification of the number of colors, and to improve the calculation accuracy of the number of colors. Therefore, by improving the calculation accuracy of the number of colors, it is possible to improve the determination accuracy of the compression format according to the number of colors, reduce the color with a high compression rate, and high image quality.

According to the invention described in claims 2 , 8 , and 11 , the same effects as in claims 1 , 7, and 10 can be obtained, and compression is performed based on the number of colors selected based on the evaluation result of the image quality evaluation. Since the method can be determined, erroneous determination of the compression format can be prevented.

According to the invention described in claims 3 , 9 , and 12 , the same effect as in claims 2 , 8, and 11 can be obtained, and when the number of colors is equal to or greater than a preset number of colors, A compression method can be determined based on a comparison result of data sizes after compression by a plurality of compression methods.

According to the fourth aspect of the present invention, it is possible to obtain the same effect as any one of the first to third aspects, as well as the mean square error (RMSE), the ratio between the image signal and the noise (PSNR). ) Can be used to objectively evaluate the image quality.

  According to the fifth aspect of the present invention, it is possible to obtain the same image quality as that of any one of the first to third aspects, but also based on the error image and the human visual system (HVS) characteristic. Can be objectively evaluated.

  According to the sixth aspect of the present invention, it is possible to obtain the same effect as any one of the first to third aspects, as well as the image data of the area where the index image data is generated and the index of the area. The image quality can be objectively evaluated based on the sharpness of the edge with the image data.

1 is a functional configuration diagram of an image forming apparatus. It is a main flowchart of an image compression process. It is a flowchart of a color number selection process. It is a flowchart of a compression method determination process. It is a flowchart of a compression system addition determination process. It is a flowchart of other color number selection processing.

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

  The image forming apparatus according to the present embodiment reads an image from a document, forms a scanned image on a sheet, and outputs it. Also, page data including image data from an external device or the like and image forming conditions for each image data. And a print function for forming an image on a sheet based on the received job and outputting the image.

FIG. 1 is a functional configuration diagram of the image forming apparatus 1.
As shown in FIG. 1, the image forming apparatus 1 includes a control unit 10, a nonvolatile memory 11, a RAM (Random Access Memory) 12, an image memory 13, an image processing unit 14, an image reading unit 20, and an operation. A display unit 30, a print unit 40, a communication unit 50, and the like are provided, and each unit is controlled by the control unit 10.

The control unit 10 includes a CPU (Central Processing Unit) and the like, reads a system program stored in the nonvolatile memory 11 and a specified program from various application programs, develops the program in the RAM 12, and develops the program in the RAM 12. Various processes are executed in cooperation with the program, and each part of the image forming apparatus 1 is centrally controlled.
For example, according to an instruction signal input from the operation display unit 30, the copy mode, the printer mode, and the scanner mode are switched, and control such as copying, printing, and reading of image data is performed.

  Further, the control unit 10 reads out the image compression processing program and various data according to the present embodiment from the nonvolatile memory 11, expands them in the RAM 12, and executes the image compression processing by the image processing unit 14 in cooperation with the program and data. Let

  In the image compression process, the image data input from the image reading unit 20 is compressed to generate a compound document file, and the compressed image data (compound document file) is written in the image memory 13 and temporarily stored. .

  In addition to various processing programs and data related to image formation, the nonvolatile memory 11 stores an image compression processing program according to the present embodiment, data processed by the various programs, and the like.

  The RAM 12 forms a work area for temporarily storing various programs executed by the control unit 10 and various data related to these programs, and stores setting information received by the operation display unit 30.

  The image memory 13 is composed of a nonvolatile readable / writable storage medium such as DRAM (Dynamic RAM) or HDD (Hard Disk Drive), and stores data processed by the image processing unit 14.

  The image processing unit 14 includes an image preprocessing unit 14a, an area forming unit 14b, an area determining unit 14c, a compression unit 14d, and a compound document file creating unit 14e.

  In the present embodiment, the image preprocessing unit 14a, the region forming unit 14b, the region determination unit 14c, the compression unit 14d, and the compound document file creation unit 14e of the image processing unit 14 are executed by the control unit 10 executing an image compression processing program. However, the present invention is not limited to this. Each unit may be configured by hardware such as an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), and a large scale integration (LSI), and the image processing unit 14 may be configured as an image processing apparatus. Further, a part of the image processing unit 14 may be realized as a hardware configuration, and the other part may be realized as software executed by the control unit 10.

  In the image preprocessing unit 14a, preprocessing is performed on the image data input from the image reading unit 20. Pre-processing includes, for example, resolution conversion, background removal processing, and the like, and increases the accuracy of processing executed by the subsequent area forming unit 14b.

  In the area forming unit 14b, the image data preprocessed by the image preprocessing unit 14a is divided into a plurality of areas and extracted. As a method for extracting the region, for example, a plurality of pixel groups including portions where pixels satisfying a predetermined condition (for example, black pixels) are continuous are extracted as one region.

  In the area discriminating unit 14c, a color number selection process in which the number of colors in each area of the image data is selected and a compression method determination process in which the compression method in each area is set based on the number of colors are executed.

  In the compression unit 14d, each area is compressed by the compression method set for each area. As a compression method, there are a reversible compression method and an irreversible compression method that can obtain a higher compression rate than the reversible compression method. Examples of the reversible compression method include MMR (Modified Modified READ) used as a compression method for character images such as characters and symbols, PNG (Portable Network Graphics) used as a compression method for line drawings such as graphics, and the like. Examples of the irreversible compression method include JPEG (Joint Photographic Experts Group) used as a compression method for background images and photographic images.

  In the compound document file creation unit 14e, the image data in which each area is compressed by the compression unit 14d is compounded, a compound document file is created, and the compound document file is stored in the image memory 13. The file format of the compound document file is, for example, PDF (Portable Document Format), XPS (XML Paper Specification), or the like.

  Further, when there is an instruction to read out image data from the control unit 10, the image processing unit 14 expands the compressed image data stored in the image memory 13 and temporarily stores it in the image memory 13. When there is an image formation instruction from the control unit 10, uncompressed image data is read in units of pages, image processing such as enlargement / reduction or direction change is performed, and the image data subjected to the image processing is output to the printing unit 40. .

  The image reading unit 20 includes an automatic document feeding unit and a reading unit, reads an image from the document, generates image data, and outputs the image data to the image processing unit 14.

  The operation display unit 30 includes a display unit including an LCD (Liquid Crystal Display), an EL (Electro-Luminescence) display, and the like, a touch panel provided to cover the display unit, and various operation key groups such as a numeric keypad. The operation display unit 30 displays various screens for inputting various setting conditions, various processing results, and the like on the display unit in accordance with a display signal input from the control unit 10. The operation display unit 30 also outputs operation signals input from various switches, buttons, numeric keys, operation key groups, or a touch panel to the control unit 10.

  The printing unit 40 includes a paper feeding unit, a paper feeding conveyance unit, an image forming unit, a fixing unit, a carry-out unit, and the like, and performs electrophotographic image forming processing based on input print data. is there.

  The communication unit 50 manages and controls data communication with an external device (not shown) connected to the network. For example, the communication unit 50 receives data to be printed from the external device, and Data is transmitted to the control unit 10 as a job, or a compound document file stored in the image memory 13 is transmitted to an external device.

Next, the operation of the present embodiment will be described.
FIG. 2 shows a main flowchart of image compression processing in the present embodiment.
The image compression processing in the present embodiment is realized as an operation of the image processing unit 14 by the control unit 10 executing an image compression processing program.

  The control unit 10 causes the image reading unit 20 to generate image data from the document, and causes the image processing unit 14 to output the image data (step S1). Examples of the file format of the generated image data include TIFF (Tagged Image File Format), JPEG, and BMP (Bit MaP).

  The control unit 10 causes the image preprocessing unit 14a to function by performing preprocessing on the image data input from the image reading unit 20 (step S2). Then, the control unit 10 divides and extracts the preprocessed image data into a plurality of regions, thereby causing the region forming unit 14b to function as a region forming unit (step S3). Identification information is added to each extracted region.

  The control unit 10 causes the region determination unit 14c to function by executing a color number selection process (step S4) described later for each region. In the color number calculation process, the number of colors in each region is selected.

  The control part 10 functions the area | region discrimination | determination part 14c by performing the compression system determination process (step S5) mentioned later after step S4. In the compression method determination process, the compression method of each area is determined based on the number of colors selected by the color number selection process, and the area determination unit 14c functions as a compression method determination unit.

  The processing result information of each area obtained by executing Steps S4 and S5 includes the number of colors, representative color values, compression method, index image data, and the like.

  Based on the processing result information of each area obtained by the area discriminating section 14c, the control section 10 compresses each area by a set compression method, and the compressed image data of each area is a compound document file creation section 14e. The compression unit 14d is caused to function (step S6).

  The control unit 10 creates a composite document file by combining the compressed image data of each area with the position information, color information (number of colors, representative color values, etc.), transparency information, etc. of the image data of each area. (Step S7), the created compound document file is stored in the image memory 13 (Step S8), and the compound document file creation unit 14e is caused to function.

FIG. 3 shows a flowchart of the color number selection process executed in step S4.
The process shown in FIG. 3 is executed for each area.

The control unit 10 calculates the number of colors based on the image data of the area targeted for the color number selection process, and causes the area determination unit 14c to function as a color number calculation unit. Then, the control unit 10 sets the calculated number of colors as the number of colors n of the area to be processed (step S11).
In step S11, a histogram of the image data of the area is created, and the number of colors is calculated using a known technique such as analysis processing of the histogram and clustering of a color set.

The control unit 10 selects the representative color value of the region to be processed based on the set number of colors n, and causes the region determination unit 14c to function as a representative color value selection unit (step S12).
In step S12, for example, a representative color value is selected using a representative color value calculation algorithm described in (Reference Document 1) and (Reference Document 2) shown below.
The selection of the representative color value is not limited to the above, and various changes can be made without departing from the gist of the present embodiment, and the concept of a suitable representative color value calculation algorithm that can be adopted is color generation. Regardless of the frequency, it is an algorithm that can calculate the color if it is important for the human visual system (HVS).

(Reference 1)
P. Heckert, “Color Image quantization for frame buffer display”, Computer graphics, Vol. 16, No. 3, p297-304, 1980
(Reference 2)
Xiaolin Wu “Color quantization by dynamic programming and principal analysis” ACM Transaction on Graphics Volume 11, Issue 4 (October 1992) Pages: 348-372

  The control unit 10 generates index image data in which each color value of the region (region to be processed) in which the representative color value is selected in step S12 is replaced with the representative color value selected in step S12, and region determination The unit 14c is caused to function as an index image generation unit (step S13).

The control unit 10 performs image quality evaluation of the image data of the region (processing target region) where the index image data is generated in step S13 and the index image data of the region, and the evaluation result is a preset evaluation result. And the region discriminating unit 14c is caused to function as an image quality evaluation color number selecting unit (step S14).
The preset evaluation result serves as a reference for determining whether or not the image quality can be maintained.

In the image quality evaluation performed in step S14, for example, the root mean square error (RMSE) between the image data of the area where the index image data is generated and the index image data of the area, the image signal and the noise are included. Any one of a ratio (PSNR; Peak Signal to Noise Ratio) and a structural similarity (SSIM) is calculated, and image quality evaluation is executed based on the calculated value.
In the present embodiment, an RMSE value is calculated, and the RMSE value is less than 10 as a preset evaluation result.

The image quality evaluation performed in step S14 is not limited to the above, and other techniques may be used in consideration of the balance between the processing speed and processing capacity and the accuracy of image quality determination.
For example, the human visual system (HVS) is known to be sensitive to distortion of a significant portion of an image and the characteristic that it is difficult to detect distortion with high spatial frequency and is easy to detect distortion with low spatial frequency. ing. Therefore, an error image between the image data of the area where the index image data is generated and the index image data of the area may be calculated, and the image quality evaluation may be performed based on the error image and the characteristics of the HVS.

  The edge indicates the characteristics of the type of image such as a character image or a line drawing such as a figure. Therefore, as another image quality evaluation executed in step S14, an edge between the image data of the area where the index image data is generated and the index image data of the area is extracted, and the edge sharpness comparison or the edge is formed. The image quality evaluation may be executed based on the comparison result of the contour shapes of the images.

  In step S14, when the evaluation result cannot maintain the image quality, that is, when the evaluation result does not correspond to the preset evaluation result (RMSE value <10) (step S14; NO), the control unit 10 is set. It is determined whether or not the number of existing colors n is greater than a preset color number threshold (step S15). The color number threshold value used as a discrimination criterion in step S15 is, for example, the number of colors discriminated as a photographic image.

  When the set color number n is equal to or smaller than a preset color number threshold (step S15; NO), the control unit 10 adds Δn to the set color number n to increase the color number ( In step S16), the process proceeds to step S12, the representative color value for the number of colors n increased in step S16 is selected again, and the process proceeds to step S13 and subsequent steps.

  The control unit 10 executes step S14; NO, step S15; NO, and step S16, thereby causing the region determination unit 14c to increase the number of colors in the region where the evaluation result does not correspond to the preset evaluation result. It functions as an increase means.

  When the set number of colors n is larger than a preset number-of-colors threshold (step S15; YES), the control unit 10 sets the compression method of the area to be processed as a photographic image compression method (for example, JPEG) is set (step S17), the compression method set in step S17 is stored in the nonvolatile memory in association with the identification information of the area to be processed, and the process is terminated.

  In step S14, when the evaluation result can maintain the image quality, that is, when the evaluation result corresponds to a preset evaluation result (RMSE value <10) (step S14; YES), the control unit 10 indicates that the evaluation result is The number n of colors set when the evaluation result is set in advance is selected, the number n of colors, the representative color value selected based on the number of colors, and the generated index image data Then, the information is stored in the nonvolatile memory in association with the identification information of the area to be processed (step S18), and this process ends.

FIG. 4 shows a flowchart of the compression method determination process executed in step S5.
This process is executed for each area for which no compression method is set.

  The control unit 10 selects any one of the plurality of areas as a processing target area of the compression method determination process, reads out the number of colors of the selected area from the nonvolatile memory, and determines the number of colors (step) S21).

When the number of colors is 2 (step S21; number of colors = 2), the control unit 10 separates the image data of the region to be processed into foreground image data and background image data (step S22).
The foreground image data is image data that forms a character image such as a character or a symbol, and the background image data is image data having a predetermined density that becomes the background of the foreground image data.

  The control unit 10 sets the reversible compression method for the foreground image data and sets the irreversible compression method for the background image data (step S23), and the setting information identifies the area to be processed. Are stored in the non-volatile memory in association with each other, and the process is terminated.

  In step S23, since the number of colors is 2, it can be determined that the image of the region to be processed is composed of one color character image (foreground image) and one color background image. Accordingly, in step S23, for example, foreground image data is preferably set as MMR used as a compression method for character images, and background image data is set as JPEG used as a compression method for photographic images.

  When the number of colors is 3 or more and 8 or less (step S21; 3 ≦ number of colors ≦ 8), the control unit 10 separates the image data of the region to be processed into a plurality of foreground image data and background image data. (Step S24).

  The control unit 10 sets a reversible compression method for a plurality of foreground image data, sets an irreversible compression method for background image data (step S25), and the setting is a processing target. The information is stored in the non-volatile memory in association with the area identification information, and the process is terminated.

  In step S25, since the number of colors is 3 or more and 8 or less, it is determined that the image of the region to be processed is composed of a character image (foreground image) of a plurality of colors and a background image of one color. it can. Therefore, in step S25, for example, it is preferable to set foreground image data of each color as MMR used as a compression method for character images, and background image data as JPEG used as a compression method for photographic images.

  When the number of colors is 9 or more and 16 or less (step S21; 9 ≦ color number ≦ 16), the control unit 10 sets a reversible compression method for the image data in the region to be processed (step S26). The setting is stored in the non-volatile memory in association with the identification information of the area to be processed, and this process is terminated.

  In step S26, since the number of colors is 9 or more and 16 or less, it can be determined that the image of the region to be processed is a line drawing such as a figure. Therefore, in step S26, for example, it is preferable to set the image data of the area to PNG used as a line drawing compression method.

  When the number of colors is 17 or more and 256 or less (step S21; 17 ≦ color number ≦ 256), the control unit 10 executes a compression method addition determination process on the image data of the region to be processed (step S27). ), The process result is stored in the nonvolatile memory in association with the identification information of the area to be processed, and the process is terminated.

  In the compression method addition determination process, when the number of colors is equal to or greater than a preset number of colors (17 in the present embodiment), the compressed data size of the processing target area is calculated for each of a plurality of different compression methods. Based on the comparison result of the data size for each compression method, the compression method of the region is determined, and the region determination unit 14c functions as a compression method addition determination unit.

FIG. 5 shows a flowchart of the compression method addition determination process executed in step S27.
The control unit 10 reads the number of colors, the representative color values, and the index image data of the region to be processed from the nonvolatile memory, and based on the number of colors, the representative color values, and the index image data, a line drawing compression method (for example, , PNG), the compressed data size (line drawing compression size P) of the area is calculated (step S31).

  In addition, the control unit 10 reads the original image data of the area to be processed from the nonvolatile memory, and compresses the data size (photograph) of the area when the photographic image compression method (for example, JPEG) is compressed. The compression size J) is calculated (step S32).

  The control unit 10 compares the line drawing compression size P calculated in step S31 with the photograph compression size J, and determines whether or not the photograph compression size J is larger than the line drawing compression size P (step S33).

  When the photograph compression size J is larger than the line drawing compression size P (step S33; YES), the control unit 10 sets a line drawing compression method for the image data of the region to be processed (step S34). This process ends.

  When the photo compression size J is equal to or smaller than the line drawing compression size P (step S33; NO), the control unit 10 sets a photo image compression method for the image data of the region to be processed (step S35). This process ends.

[Modification]
The color number selection process executed in step S4 may use the process shown in FIG. 6 instead of the process shown in FIG. FIG. 6 shows a flowchart of another color number selection process executed in step S4. The process shown in FIG. 6 is executed for each area.

The control unit 10 calculates the number of colors based on the image data of the region that is the processing target of the color number selection process, and causes the region forming unit 14b to function as a color number calculation unit (step S41).
In step S41, a histogram of the image data of the area is created, and the number of colors is calculated using a known technique such as analysis processing of the histogram and clustering of a color set.

  Based on the number of colors calculated in step S41, the control unit 10 generates a plurality of color number candidates (first to N color number candidates) for the region, and causes the region determination unit 14c to function as a color number candidate generation unit. (Step S42).

  In step S42, using the number of colors calculated in step S41 as a reference, the number of colors before and after the number of colors, the number of colors smaller than the number of colors, or the number of colors larger than the number of colors. The number of colors is generated, and the generated number of colors and the reference number of colors are generated as color number candidates. For example, if the number of colors calculated in step S41 is 10, the two preceding and following numbers of colors (8, 9, 11, 12) are generated with reference to 10, and 8, 9, 10, 11, 12 are the first. Generated as 1 to 5 color candidates.

The control unit 10 performs the processes of steps S43 to S45 for each color number candidate.
Processing for the first color number candidate will be described as a representative of processing for other color number candidates.
Based on the number of colors generated as the first color number candidate, the control unit 10 selects the representative color value of the region to be processed, and causes the region determination unit 14c to function as a representative color value selection unit (step S43). ). Since the selection of the representative color value in step S43 is the same as that in step S12, description thereof is omitted.

  The control unit 10 generates index image data in which each color value of the region to be processed is replaced with the representative color value selected in step S43, and causes the region determination unit 14c to function as an index image generation unit (step S44). ).

  The control unit 10 evaluates the image quality of the image data of the region to be processed and the index image data of the region, and causes the region determination unit 14c to function as an image quality evaluation unit (step S45). Since the image quality evaluation executed in step S45 is the same as that in step S14, description thereof is omitted. In the present embodiment, the RMSE value is calculated.

The control unit 10 selects any one of the color number candidates from the plurality of color number candidates based on the image quality evaluation result of each color number candidate, and causes the region determination unit 14c to function as the color number determination unit. For example, the candidate for the number of colors from which the result of the image quality evaluation corresponding to the preset evaluation result is obtained is selected, and the candidate for the number of colors having the smallest number of colors is selected from the selected candidate number of colors. Then, the control unit 10 determines the selected color number candidate as the number of colors in the region to be processed.
Then, the control unit 10 associates the determined color number n, the representative color value selected based on the color number, and the generated index image data with the identification information of the area to be processed. Is stored in the non-volatile memory (step S46), and this process is terminated.

  In the present embodiment, the processing of each step illustrated in FIGS. 3 to 6 has been described as being realized by the control unit 10 executing the image processing program. However, the present invention is not limited thereto, and the processing of each step is performed. ASIC (Application Specific Integrated Circuit), FPGA (Field Programmable Gate Array), LSI (Large Scale Integration), or other hardware, or a part of the hardware configuration is executed by the controller 10. It may be realized by a program.

As described above, according to the present embodiment, objective image quality evaluation is performed on image data and index image data of a region to be processed, and the evaluation result corresponds to a preset evaluation result. In this case, the calculated number of colors is selected, and if the evaluation result does not correspond to the preset evaluation result, the number of colors is increased, the image quality is evaluated again, and the color is selected according to the evaluation result. You can select a number. Alternatively, one of a plurality of color number candidates can be selected based on the evaluation result by the image quality evaluation unit for each color number candidate.
Therefore, the number of colors can be selected based on the image quality evaluation, and it is possible to prevent under- and over-identification of the number of colors when the number of colors is calculated using a frequency threshold such as a histogram, thereby improving the calculation accuracy of the number of colors. Therefore, by improving the calculation accuracy of the number of colors, it is possible to improve the determination accuracy of the compression format according to the number of colors, reduce the color with a high compression rate, and high image quality.

In addition, since the compression method can be determined based on the number of colors selected based on the evaluation result of the image quality evaluation, erroneous determination of the compression format can be prevented. That is, image data in a region with a small number of colors forming a character image or line drawing can be compressed with a compression method that maintains high image quality, and image data in a region with a large number of colors forming a photographic image or the like. Can be compressed at a high compression rate.
In particular, when the number of colors is greater than or equal to a preset number of colors (17 in this embodiment), the compression method is based on the comparison results of the data sizes after compression by a plurality of compression methods (PNG, JPEG). And the compression efficiency of the image data can be improved.

  Also, an image quality evaluation method based on any one of calculated values of mean square error (RMSE), image signal to noise ratio (PSNR), and structural similarity (SSIM), error image and human visual system (HVS) characteristics The image quality can be objectively evaluated by the image quality evaluation method based on the above and the image quality evaluation method based on the edge sharpness between the image data of the area where the index image data is generated and the index image data of the area.

In the above description, the example in which the nonvolatile memory 11 is used as the computer-readable medium of the program according to the present invention has been disclosed, but the present invention is not limited to this example.
Other computer-readable media include magnetic tape, magnetic disks such as FD (Floppy (registered trademark) Disk), DVD (Digital Versatile Disk), CD-ROM (Compact Disk Read Only Memo), CD-R (Compact Disk). It is possible to apply an optical recording medium such as Recordable) or MO (Magneto-Optical disk), or a portable recording medium such as a flash memory.
Further, a carrier wave (carrier wave) is also applied to the present invention as a medium for providing program data according to the present invention via a communication line.

Further, the program according to the present invention need not include all modules for causing a computer to execute the processing described above. For example, the processing of the present invention may be executed using various general-purpose programs that can be separately installed in a computer, such as a communication program and a program included in an OS (Operating System).
Therefore, it is not always necessary to record all modules of the program according to the present invention on a recording medium, and it is not always necessary to transmit all modules. Furthermore, the predetermined processing may be executed using dedicated hardware.

  Further, the present invention is not limited to the contents of the above embodiment, and can be appropriately changed without departing from the gist of the present invention.

DESCRIPTION OF SYMBOLS 1 Image forming apparatus 10 Control part 11 Non-volatile memory 12 RAM
13 Image memory 14 Image processing unit 14a Image preprocessing unit 14b Region forming unit 14c Region determining unit 14d Compression unit 14e Compound document file creating unit 20 Image reading unit 30 Operation display unit 40 Print unit 50 Communication unit

Claims (12)

An area forming unit that divides the image data into a plurality of areas based on a predetermined condition ;
A color number calculation unit for calculating a first color number for each of the regions;
A representative color value selection unit that selects a representative color value for each region based on the first number of colors calculated by the color number calculation unit;
An index image generation unit that generates index image data in which each color value of the region where the representative color value is selected is replaced with the representative color value;
Image quality evaluation is performed between the image data of the area where the index image data is generated and the index image data of the area, and the number of colors calculated when the evaluation result corresponds to a preset evaluation result is selected. An image quality evaluation color selection section;
A color number increasing unit for calculating a second color number obtained by adding a predetermined number to the first color number for an area where the evaluation result by the image quality evaluation color number selecting unit does not correspond to a preset evaluation result;
Equipped with a,
The representative color value selection unit, for an area where the evaluation result does not correspond to a preset evaluation result, based on the second color number calculated by the color number increasing unit, the representative color value of the area Image processing device that elects again . A compression method determination unit that determines a compression method of the region based on the number of colors selected by the image quality evaluation color number selection unit;
An image processing apparatus according to claim 1. The compression method determination unit
When the number of colors in the area is greater than or equal to a preset number of colors, the compressed data size of the area is calculated for each of a plurality of different compression methods, and the data size comparison result for each compression method is used. A compression method addition determination unit that determines the compression method of the area,
The image processing apparatus according to claim 2. The image quality evaluation color number selection unit is
Calculating one of a mean square error between the image data of the area where the index image data is generated and the index image data of the area, and the ratio of the image signal and noise, and evaluating the image quality based on the calculated value;
The image processing apparatus according to claim 1. The image quality evaluation color number selection unit is
Calculating an error image between the image data of the area where the index image data is generated and the index image data of the area, and evaluating the image quality based on the error image and human visual system characteristics;
The image processing apparatus according to claim 1. The image quality evaluation color number selection unit is
The image quality is evaluated based on the sharpness of the edge between the image data of the area where the index image data is generated and the index image data of the area.
The image processing apparatus according to claim 1. A region forming step of dividing the image data into a plurality of regions based on a predetermined condition ;
A color number calculating step of calculating a first color number for each of the regions;
A representative color value selection step of selecting a representative color value for each region based on the first number of colors calculated in the color number calculation step;
An index image generation step of generating index image data in which each color value of the region where the representative color value is selected is replaced with the representative color value;
Image quality evaluation is performed between the image data of the area where the index image data is generated and the index image data of the area, and the number of colors calculated when the evaluation result corresponds to a preset evaluation result is selected. Image quality evaluation color selection process,
A color number increasing step of calculating a second color number obtained by adding a predetermined number to the first color number for an area where the evaluation result by the image quality evaluation color number selecting step does not correspond to a preset evaluation result;
For an area where the evaluation result does not correspond to a preset evaluation result, based on the second color number calculated by the color number increasing step, the representative color value re-selecting the representative color value of the area is selected again. A selection process;
An image processing method including: A compression method determination step of determining a compression method of the region based on the number of colors selected in the image quality evaluation color number selection step;
The image processing method according to claim 7. The compression method determination step includes
When the number of colors in the area is greater than or equal to a preset number of colors, the compressed data size of the area is calculated for each of a plurality of different compression methods, and the data size comparison result for each compression method is used. Including a compression method addition determination step of determining the compression method of the area,
The image processing method according to claim 8. Computer
Area forming means for dividing the image data into a plurality of areas based on a predetermined condition ;
A color number calculating means for calculating a first color number for each of the areas;
Representative color value selection means for selecting a representative color value for each region based on the first color number calculated by the color number calculation means;
Index image generating means for generating index image data in which each color value of the area where the representative color value is selected is replaced with the representative color value;
Image quality evaluation is performed between the image data of the area where the index image data is generated and the index image data of the area, and the number of colors calculated when the evaluation result corresponds to a preset evaluation result is selected. Image quality evaluation color selection means,
Color number increasing means for calculating a second color number obtained by adding a predetermined number to the first color number for an area where the evaluation result by the image quality evaluation color number selecting means does not correspond to a preset evaluation result;
Is a program for functioning as
The representative color value selection means, for the area where the evaluation result does not correspond to a preset evaluation result, based on the second color number calculated by the color number increasing means, the representative color value of the area Program to elect again . The computer,
Based on the number of colors selected by the image quality evaluation color number selection means, it functions as a compression method determination means for determining the compression method of the area.
The program according to claim 10. The compression method determination means includes
When the number of colors in the area is greater than or equal to a preset number of colors, the compressed data size of the area is calculated for each of a plurality of different compression methods, and the data size comparison result for each compression method is used. And a compression method addition determination means for determining the compression method of the area,
The program according to claim 11.

Images