JP7205689B2 - Image processing device and computer program - Google Patents

Description

The present specification relates to image processing for image data generated by reading printed matter using an image sensor, and more particularly to image processing for identifying character pixels in an image.

In image data generated by reading printed matter using an image sensor, halftone dots included in the printed matter appear in the image indicated by the image data. Pixels forming such halftone dots are likely to be erroneously identified as character pixels when identifying character pixels in an image.

The image processing apparatus disclosed in Japanese Patent Application Laid-Open No. 2002-200013 performs edge determination for determining whether each pixel is an edge and halftone dot determination for determining whether each pixel is a halftone dot. The image processing device identifies pixels that are edges and are not halftone dots as pixels representing characters.

JP-A-4-318659 JP-A-3-208467

As described above, even in an image including halftone dots, for example, there is a technique that prevents erroneous identification of character pixels in an area without characters due to halftone dots, thereby accurately identifying character pixels. was wanted.

This specification discloses a technique that can accurately identify character pixels in a target image even if the image includes halftone dots.

The technology disclosed in this specification has been made to solve at least part of the above-described problems, and can be implemented as the following application examples.

[Application Example 1] An image processing apparatus comprising: an image acquisition unit that acquires target image data generated by reading printed matter using an image sensor; a first identifying unit that identifies a plurality of candidate pixels that are candidates for character pixels from among a plurality of pixels in a target image that is captured, wherein a plurality of edge pixels forming an edge in the target image are identified; and a placement unit for placing a plurality of blocks in the target image, wherein the plurality of blocks is a specific block indicating characteristics of a background. arranged in the target image using the placement unit and the target image data, including a plurality of background blocks having feature values and non-background blocks having feature values different from the specific feature values; a continuation number identifying unit that identifies a continuation number, which is the number of continuations of non-background blocks among a plurality of blocks in the background block; and a second identification unit that identifies a character block from the plurality of blocks based on the continuation number. and a third identifying unit that identifies the plurality of character pixels in the target image based on the result of identifying the candidate pixel and the result of identifying the character block.

Objects that are different from the background have different feature continuity depending on the type, such as characters or photographs. According to the above configuration, character blocks are identified based on the number of consecutive non-background blocks that are different from the background, so character blocks can be appropriately identified in consideration of the continuity of the features described above. Therefore, the character pixels in the target image can be accurately identified based on the identification result of the candidate pixel and the identification result of the character block.

It should be noted that the technology disclosed in this specification can be implemented in various forms. It can be implemented in the form of a program, a recording medium recording the computer program, or the like.

1 is a block diagram showing the configuration of a multi-function peripheral 200, which is an example of an image processing apparatus; FIG. 4 is a flowchart of image processing; FIG. 4 is a first diagram showing an example of an image used in image processing; It is a figure which shows an example of the image used by character specific processing. 4 is a flowchart of first binary image data generation processing; 6 is a flowchart of edge enhancement processing; FIG. 4 is an explanatory diagram of smoothed R component image data and edge-enhanced R component image data; FIG. 10 is a diagram showing an example of a tone curve for level correction processing; 9 is a flowchart of second binary image data generation processing; FIG. 4 is an explanatory diagram of minimum component values and maximum component values of scan data; FIG. 4 is a second diagram showing an example of an image used for image processing; 4 is a flowchart of block determination processing of the first embodiment; 4 is a flowchart of block determination processing of the first embodiment; FIG. 3 is an explanatory diagram of a plurality of blocks BL arranged on a scanned image SI; FIG. 4 is an explanatory diagram of run length information; FIG. 4 is an explanatory diagram of various blocks specified on the scanned image SI; 10 is a flowchart of block determination processing of the second embodiment;

A. Example:
A-1: Configuration of MFP 200 Embodiments will be described based on examples. FIG. 1 is a block diagram showing the configuration of a multifunction machine 200, which is an example of an image processing apparatus. The MFP 200 includes a CPU 210 that is a processor that controls an image processing device, a volatile storage device 220 such as a DRAM, a nonvolatile storage device 230 such as a flash memory or a hard disk drive, a display unit 240 such as a liquid crystal display, An operation unit 250 including a touch panel and buttons superimposed on a liquid crystal display, an interface (communication IF) 270 for communicating with an external device such as the user's terminal device 100, a print execution unit 280, and a read execution unit 290. , is equipped with

Reading execution unit 290 generates scan data by optically reading a document using a one-dimensional image sensor under the control of CPU 210 . Under the control of the CPU 210, the print execution unit 280 uses a plurality of types of toner, specifically, cyan (C), magenta (M), yellow (Y), and black (K) toners, as colorants. An image is printed on a print medium such as paper using a laser method. Specifically, the print execution unit 280 exposes the photosensitive drum to form an electrostatic latent image, and adheres toner to the electrostatic latent image to form a toner image. The print execution unit 280 transfers the toner image formed on the photosensitive drum onto the paper. In a modified example, the print execution unit 280 may be an inkjet print execution unit that forms an image on paper by ejecting ink as a coloring material.

Volatile storage device 220 provides a buffer area for temporarily storing various intermediate data generated when CPU 210 performs processing. A computer program PG is stored in the nonvolatile storage device 230 . The computer program PG is a control program that causes the CPU 210 to control the MFP 200 . In this embodiment, the computer program PG is provided in a form pre-stored in the non-volatile storage device 230 when the MFP 200 is manufactured. Alternatively, the computer program PG may be provided in the form of being downloaded from a server, or may be provided in the form of being stored on a DVD-ROM or the like. The CPU 210 can execute image processing, which will be described later, by executing the computer program PG.

A-2: Image Processing FIG. 2 is a flowchart of image processing. This image processing is executed, for example, when the user places a document on the document platen of the reading execution unit 290 and inputs a copy execution instruction. This image processing acquires scan data generated by reading a document using the reading execution unit 290, and uses the scan data to generate print data representing the document, thereby making a so-called copy of the document. This is the process to be implemented.

In S<b>10 , the CPU 210 reads the document placed on the platen by the user using the reading execution unit 290 to generate scan data as target image data. The document is, for example, a printed matter in which an image is printed by the MFP 200 or a printer (not shown). The generated scan data is stored in a buffer area of volatile storage device 220 (FIG. 1). The scan data includes a plurality of pixel values, and each of the plurality of pixel values represents the color of the pixel with a color value (also called an RGB value) in the RGB color system. That is, the scan data is RGB image data. The RGB value of one pixel includes, for example, three color component values of red (R), green (G), and blue (B) (hereinafter also referred to as R value, G value, and B value). I'm in. In this embodiment, the number of gradations of each component value is 256 gradations.

Scan data, which is RGB image data, includes three component image data (R component image data, G component image data, and B component image data) corresponding to three color components forming the RGB color system. can be said. Each component image data is image data in which the value of one kind of color component is used as the pixel value.

FIG. 3 is a first diagram showing an example of an image used in image processing. FIG. 3A shows an example of a scanned image SI represented by scan data. Scan image SI includes a plurality of pixels. The plurality of pixels are arranged in a matrix along a first direction D1 and a second direction D2 orthogonal to the first direction D1.

The scanned image SI of FIG. 3A includes a white background Bg1 indicating the background color of the paper of the original, three objects Ob1 to Ob3 different from characters, two characters Ob4 and Ob5, and two characters Ob4 and Ob5. and a background Bg2 of characters Ob4 and Ob5. An object different from text is, for example, a photograph. Background Bg2 is a uniform image having a color different from white.

In S20, the CPU 210 executes character identification processing on the scan data. The character identification process identifies character pixels by classifying a plurality of pixels in the scan image SI into a plurality of character pixels indicating characters and a plurality of non-character pixels not indicating characters. is.

By the character identification process, for example, binary image data (also referred to as character identification data) is generated in which the value of character pixels is set to "1" and the value of non-character pixels is set to "0". FIG. 3B shows an example of the character identification image TI indicated by the character identification data. In this character identification image TI, a plurality of pixels forming edges of two characters Ob4 and Ob5 in the scanned image SI are identified as character pixels Tp4 and Tp5. For relatively large characters, the pixels forming the edges of the characters are identified as character pixels, and for relatively small characters, the entire pixels forming the characters are identified as character pixels. Details of the character identification process will be described later.

In S30, the CPU 210 executes halftone dot smoothing processing on the scan data to generate smoothed image data representing a smoothed image. Specifically, the CPU 210 executes smoothing processing using a smoothing filter such as a Gaussian filter on each of the values of a plurality of non-character pixels included in the scan data, and performs smoothing processing. Calculate values for a plurality of non-character pixels. Non-character pixels to be smoothed are specified by referring to the character specifying data generated by the classification process of S20. The CPU 210 generates smoothed image data including the values of a plurality of character pixels included in the scan data and the values of a plurality of non-character pixels that have been smoothed.

FIG. 3C shows the smoothed image GI represented by the smoothed image data. The smoothed image GI includes a white background Bg1g, objects Ob1 to Ob5 in the scanned image SI, objects Ob1g to Ob5g obtained by smoothing the background Bg2, and a background Bg2g. Of these objects Ob1g to Ob5g and background Bg2g, portions other than the characters Ob4g and Ob5g (also referred to as non-character portions) are smoothed compared to the scanned image SI.

In S40, the CPU 210 executes character sharpening processing on the smoothed image data to generate processed image data. Specifically, the CPU 210 executes sharpening processing such as unsharp mask processing and processing applying a sharpening filter to each of the values of a plurality of character pixels included in the smoothed image data. Values of a plurality of sharpened character pixels are calculated. The character pixels to be sharpened are specified by referring to the character specifying data generated by the classification process of S20. Then, the CPU 210 calculates the values of the plurality of non-character pixels (the values of the plurality of non-character pixels that have undergone the smoothing process) and the values of the plurality of character pixels that have undergone the sharpening process, which are included in the smoothed image data. , to generate processed image data including . The values of the plurality of character pixels included in the smoothed image data are the same as the values of the plurality of character pixels included in the scan data because they are not subject to smoothing processing. Therefore, it can be said that the character sharpening process of this step is performed on the values of a plurality of character pixels included in the scan data.

FIG. 3D shows the processed image FI indicated by the processed image data. The processed image FI includes a white background Bg1f, objects Ob1 to Ob5 in the scanned image SI, objects Ob1f to Ob5f corresponding to the background Bg2, and a background Bg2f. Among these objects Ob1f to Ob5f and the background Bg2f, the edges of the characters Ob4f and Ob5f are sharpened compared to the characters Ob4 and Ob5 in the scanned image SI and the characters Ob4g and Ob5g in the smoothed image GI. there is Edges of objects Ob1f to Ob3f other than characters and background Bg2f are not sharpened.

As can be seen from the above description, the objects Ob1f to Ob5f and the background Bg2f in the processed image FI include sharpened characters and smoothed non-characters.

In S50, the CPU 210 executes print data generation processing for generating print data using the processed image data. Specifically, color conversion processing is performed on the processed image data, which is RGB image data, to obtain a color having color components (C, M, Y, and K components) corresponding to the color materials used for printing. CMYK image data is generated that indicates the color of each pixel with the CMYK values. Color conversion processing is executed, for example, with reference to a known lookup table. Halftone processing is performed on the CMYK value image data to generate dot data indicating the state of dot formation for each color material used for printing and for each pixel. The dot formation state can take, for example, two types of dot states and no dot states, and four types of states of large dots, medium dots, small dots, and no dots. Halftone processing is performed, for example, according to a dither method or an error diffusion method. The dot data are rearranged in order of use at the time of printing, and print data is generated by adding a print command to the dot data.

In S60, the CPU 210 executes print processing and ends the image processing. Specifically, CPU 210 supplies print data to print execution unit 280 and causes print execution unit 280 to print the processed image.

According to the image processing described above, the CPU 210 as the image processing unit performs the first image processing (specifically, the edge sharpening processing) on the values of the specified plurality of character pixels in the scan data. ) is executed (S40), and second image processing different from the first image processing (specifically, dot smoothing processing) is executed on the values of a plurality of non-character pixels (S30). , to generate the processed image data. As a result, different image processing is performed on the value of the character pixel and the value of the pixel different from the character pixel, so appropriate image processing for the scan data can be realized. Note that in a modified example, the character sharpening process of S40 may be performed first, and then the halftone dot smoothing process of S30 may be performed.

More specifically, processed image data including a plurality of sharpened text pixel values and a plurality of smoothed non-text pixel values is generated (S30, S40). . As a result, processed image data representing a processed image FI with a good appearance can be generated.

For example, as shown in the processed image FI of FIG. 3D, in the processed image data, sharpened values are used as the values of the character pixels. As a result, the characters in the processed image FI look sharp, so that, for example, the appearance of the printed processed image FI can be improved.

In the processed image data, smoothed values are used for the values of the background Bg2 in the processed image FI and the values of the non-character pixels that constitute objects such as photographs that are different from characters. As a result, it is possible to suppress, for example, halftone dots that cause moire from appearing in a portion of the processed image FI that is different from the characters, thereby suppressing the occurrence of defects such as moire in the processed image FI to be printed. can. As a result, the appearance of the processed image FI to be printed can be improved. Moreover, since edges in the photograph are prevented from being excessively emphasized, the appearance of the image FI can be improved.

For example, a document used to generate scan data is a printed matter on which an image is printed. Therefore, for example, a uniform portion such as the background Bg2 having a color different from white in the document forms a halftone dot when viewed at the dot level for forming an image. A halftone dot includes a plurality of dots and a portion where no dots are arranged (portion indicating the background color of the document). For this reason, when viewed at the pixel level, halftone dots are shown in the area indicating the background Bg2 in the scan image SI. Dots in halftone dots are arranged with periodicity due to the influence of a dither matrix used when printing an original. For this reason, when printing is performed using scan data, the halftone dot periodic component existing in the original image (scanned image SI) before halftone processing and the halftone dot dot composing the printed image Moire tends to appear due to interference between periodic components and . In the processed image FI of the present embodiment, the smoothing process reduces periodic components of dots in portions different from edges in the original image (scanned image SI). As a result, when the processed image FI is printed using the processed image data, for example, moiré can be suppressed from occurring in the printed processed image FI.

In particular, in the image processing described above, the CPU 210 as a print data generation unit uses the processed image data to generate print data (S50). Possible appropriate print data can be generated.

A-3: Character Identification Processing The character identification processing of S20 in FIG. 2 will be described. In S22, the CPU 210 uses the scan data to execute a first binary image data generation process to generate first binary image data. The first binary image data is binary data representing edge pixels and non-edge pixels. Here, edge pixels indicated by the first binary image data are also called first edge pixels, and non-edge pixels indicated by the first binary image data are also called first non-edge pixels. Details of the first binary image data generation process will be described later.

In S24, the CPU 210 uses the scan data to execute a second binary image data generation process to generate second binary image data. The second binary image data, like the first binary image data, is binary data representing edge pixels and non-edge pixels. The second binary image data is generated by different processing from the first binary image data, and is data different from the first binary image data. Here, edge pixels indicated by the second binary image data are also called second edge pixels, and non-edge pixels indicated by the second binary image data are also called second non-edge pixels. Details of the second binary image data generation process will be described later.

In S26, the CPU 210 executes logical sum synthesis processing for synthesizing the first binary image data generated in S22 and the second binary image data generated in S24. Binary image data (also called edge identification data) indicating edge pixels and non-edge pixels identified by . Specifically, the CPU 210 generates binary image data as edge identification data by calculating the logical sum of each pixel of the first binary image data and the second binary image data.

In other words, the CPU 210 as a first specifying unit determines a plurality of first edge pixels specified by the first binary image data and a plurality of edge pixels specified by the second binary image data. 2 edge pixels, and does not include pixels different from either the first edge pixel or the second edge pixel, is finally defined as a plurality of edge pixels (character candidate pixels). Identify. As a result, using the first binary image data and the second binary image data, it is possible to accurately determine whether a pixel in the target image is an edge pixel. For example, omission of specifying edge pixels in the scanned image SI can be effectively reduced.

An edge pixel specified by edge specifying data is a pixel that is a candidate for a character pixel, and is also called a character candidate pixel. For example, the edge identification data has a value of "1" for character candidate pixels (edge pixels in this embodiment) and a value of "0" for pixels that are not character candidate pixels (non-edge pixels in this embodiment). binary image data.

FIG. 4 is a diagram showing an example of an image used in character identification processing. FIG. 4A shows an example of an edge identification image EI indicated by edge identification data. This edge identification image EI includes a plurality of edge pixels forming edges Eg1 to Eg5 of objects Ob1 to Ob5 in the scan image SI, and a plurality of edges forming an edge Eg8 on the boundary between the background Bg1 and the background Bg2. pixels are identified as character candidate pixels. Thus, edges indicated by character candidate pixels include character edges. The edges also include edges such as thin lines included in objects other than characters (for example, photographs).

In S28, the CPU 210 executes block determination processing on the scan data to generate binary image data (block determination data (also called The block determination process uses scan data to determine whether each of a plurality of blocks arranged in the scan image SI is a character block or not. One block is a rectangular area containing N (N is an integer equal to or greater than 2) pixels. The block determination data is binary image data in which the value of pixels forming a character block is "1" and the value of pixels forming a non-character block is "0".

FIG. 4B shows an example of the block determination image BI indicated by the block determination data. Character blocks Bk4 and Bk5 indicating areas where characters Ob4 and Ob5 in the scan image SI are arranged are specified in the block determination image BI.

In S29, the CPU 210 executes a logical AND synthesizing process for synthesizing the edge specifying data generated in S26 and the block determination data generated in S28, thereby obtaining the character pixels and the non-character pixels. character identification data (see FIG. 3B). Specifically, the CPU 210 generates binary image data as character identification data by taking a logical AND of each pixel of the edge identification data and the block determination data. In other words, the CPU 210 identifies, as character pixels, pixels located in the character block identified by the block determination data among the plurality of candidate pixels identified by the edge identification data. The CPU 210 selects, among the plurality of pixels in the scan image SI, pixels that are different from the plurality of candidate pixels specified by the edge specifying data and that are located in blocks different from the character block specified by the block determination data. pixels are identified as non-character pixels. Once the character identification data is generated, the character identification process is terminated.

A-4: First Binary Image Data Generation Processing The first binary image data generation processing of S22 in FIG. 2 will be described. FIG. 5 is a flowchart of the first binary image data generation process. In S100, the CPU 210 executes smoothing processing on each of the three component image data included in the scan data, that is, the R component image data, the G component image data, and the B component image data. As a result, three pieces of smoothed component image data, that is, smoothed R component image data, smoothed G component image data, and smoothed B component image data are generated.

The smoothing process is a process of smoothing the component image indicated by the component image data to be processed. The smoothing process of this embodiment is a process of applying a predetermined smoothing filter to the value of each pixel of the component image data to be processed to calculate the value of each smoothed pixel. For the smoothing filter, for example, a Gaussian filter having a size of 7 vertical pixels×7 horizontal pixels is used.

In S110, edge enhancement processing is performed on each of the three pieces of smoothed component image data to obtain three pieces of edge-enhanced component image data, that is, edge-enhanced R component image data, edge G component image data that has been emphasized and B component image data that has been edge emphasized are generated.

FIG. 6 is a flowchart of edge enhancement processing. Here, it is assumed that smoothed R component image data is to be processed. Similar processing is performed on the smoothed G component image data and the smoothed B component image data.

In S200, the CPU 210 prepares canvas data for generating edge-enhanced R component image data in the memory (specifically, the buffer area of the volatile storage device 220). The canvas (initial image) indicated by the canvas data is an image having the same size as the scanned image SI, that is, an image having the same number of pixels. The value of each pixel of canvas data is a predetermined initial value (eg, 0).

In S205, the CPU 210 selects one target pixel from a plurality of pixels in the smoothed R component image indicated by the smoothed R component image data.

At S210, the CPU 210 calculates a mask value MV corresponding to the pixel of interest. The mask value MV is obtained by smoothing the value TV of the pixel of interest using the value TV of the pixel of interest and the values of a predetermined number of surrounding pixels including four pixels adjacent to the pixel of interest above, below, left and right. Calculated by processing. For this reason, the mask value MV is also called a smooth value. Specifically, the average value of the values of 100 pixels within a rectangular range of 10 vertical pixels×10 horizontal pixels centered on the pixel of interest is calculated as the mask value MV corresponding to the pixel of interest.

In S220, the CPU 210 calculates the difference ΔV between the value TV of the pixel of interest and the mask value MV corresponding to the pixel of interest (ΔV=(TV−MV)).

In S230, CPU 210 determines whether or not difference ΔV is equal to or greater than a reference value. Specifically, it is determined whether or not the difference ΔV is greater than or equal to a predetermined threshold TH. The threshold value TH is a value of about 20 to 30, for example, when the component value is a value of 256 gradations in a range of 0 to 255.

If the difference ΔV is equal to or greater than the reference (S230: YES), in S240 the CPU 210 converts the sum (TV+ΔV) of the value TV of the target pixel and the difference ΔV corresponding to the target pixel to the processed value. Calculate as If the difference ΔV is less than the reference (S230: NO), the CPU 210 skips S240.

At S245, the CPU 210 records the value of the target pixel in the canvas data prepared at S200. When S240 is executed, the sum of the target pixel value TV calculated in S240 and the difference ΔV corresponding to the target pixel is recorded in the canvas data as a processed value. If S240 is skipped, the value of the pixel of interest in the smoothed R component image data is recorded as is in the canvas data.

In S250, CPU 210 determines whether or not all pixels in the R component image have been processed as pixels of interest. If there is an unprocessed pixel (S250: NO), the CPU 210 returns to S205 and selects the unprocessed pixel as the pixel of interest. When all pixels have been processed (S250: YES), the CPU 210 terminates the edge enhancement process. At this point, edge-enhanced R component image data has been generated.

FIG. 7 is an explanatory diagram of smoothed R component image data and edge-enhanced R component image data. FIG. 7A shows a graph conceptually showing the R component image data before the smoothing process of S100 of FIG. FIGS. 7B and 7C show graphs conceptually showing smoothed R component image data and edge-enhanced R component image data, respectively. In each graph, the left portion conceptually shows a halftone dot area indicating halftone dots, and the right portion conceptually shows an edge area indicating the edge of an object such as a character. The vertical axis of each graph indicates the value of the R component, and the horizontal axis indicates the position in a predetermined direction (eg, the first direction D1 in FIG. 3).

The R component image data before the smoothing process includes, for example, a plurality of valleys C1 to C3 corresponding to a plurality of halftone dots and a plurality of gaps between the halftone dots in the halftone dot area, A plurality of peaks P1 and P2 appear (FIG. 7(A)). If there remains a large difference in the value of the R component between the troughs C1 to C3 and the plurality of peaks P1 and P2, the R Edge pixels representing halftone dots are likely to be identified due to the difference in component values. A halftone dot area includes halftone dots when viewed from a pixel-level viewpoint (a microscopic viewpoint where halftone dots can be recognized), but when viewed from an observer's viewpoint (a macroscopic viewpoint where halftone dots cannot be recognized) , is a uniform region. Therefore, in this embodiment, edge pixels caused by halftone dots should not be specified in the halftone dot area. This is because halftone dot areas are preferably smoothed in S30 of FIG. 2 and should not be sharpened in S40. This is because if the edges of the halftone dots are sharpened, the periodicity of the halftone dots becomes conspicuous, and moire becomes more conspicuous when the image is printed. For example, edge pixels should not be identified in uniform parts such as the background Bg2 in the scan image SI and in parts different from the edge of the object.

In the smoothed R component image data, for example, in the halftone dot area, the differences in the R component values between the plurality of valleys C1a to C3a and the plurality of peaks P1a and P2a are reduced. , is sufficiently smaller than the R component image data before smoothing (FIG. 7B).

Here, in the edge enhancement processing of this embodiment, the edge enhancement effect increases as the difference ΔV between the target pixel value TV and the mask value MV corresponding to the target pixel increases. For this reason, the effect of edge enhancement is reduced in a flat area with a relatively small difference in R component value, such as the halftone dot area in FIG. 7B. In addition, in the edge enhancement processing of this embodiment, when the difference ΔV is less than the reference, edge enhancement is not performed, and the pixel values of the smoothed R component image data are used as they are (see FIG. 6). S230). As a result, in the edge-enhanced R component image data, for example, a plurality of valleys C1b to C3b, a plurality of peaks P1b, and a plurality of peaks P1b, The difference between the R component values of P2b and P2b is not large compared to the smoothed R component image data (FIG. 7(C)). That is, in the edge-enhanced R component image data, similar to the smoothed R component image data, the values of the R components of the plurality of valleys C1b to C3b and the plurality of peaks P1b and P2b are The difference is sufficiently small (FIG. 7(C)) compared to the R component image data (FIG. 7(A)) before smoothing.

In the R component image data before the smoothing process, for example, in an edge region indicating the edge of an object such as a character, a fluctuating portion E1 in which the value of the R component changes sharply corresponding to the edge appears (see FIG. 7 ( A)). In such a fluctuation portion E1, the greater the change in value, the easier it is for the edge pixel indicating the edge of the object to be identified in the binarization processing of S150, which will be described later, due to the difference in the value of the R component.

In the R component image data that has been smoothed, for example, in the edge region, the change in value in the fluctuation portion E1a is smaller (slowly (Fig. 7(B)).

However, the change in the value in the fluctuating portion E1a corresponding to the edge of the object such as characters is sufficiently larger than the change in the value in the halftone dot area, so the edge emphasis process returns to a rapid change. As a result, in the edge region of the edge-enhanced R component image data, the change in the R component value of the fluctuation portion E1b is greater than that of the smoothed R component image data (FIG. 7 ( C)). Therefore, in the edge region of the edge-enhanced R component image data, the change in value in the fluctuation portion E1b is approximately the same as or sharper than that of the R component image data before the smoothing process. (FIG. 7(C)).

As can be seen from the above description, in this embodiment, the smoothing process (S100) and the edge enhancement process (S110) are performed on each component image data in this order, so that the edge of the halftone dot is It is possible to suppress the edge pixels indicating the edge from being specified, and to promote the edge pixels indicating the edge of the object such as characters to be specified. As a result, a plurality of edge pixels in the scanned image SI can be appropriately specified.

When three pieces of enhanced component image data corresponding to the three color components of R, G, and B are generated, at 120 in FIG. , to generate luminance image data. The brightness image data is data representing the brightness of a plurality of pixels in the enhanced image indicated by the three pieces of enhanced component image data. Specifically, the CPU 210 calculates the luminance Y of each pixel using the R value, G value, and B value of each pixel obtained from the three pieces of component image data that have undergone the enhancement process. The luminance Y is, for example, the weighted average of the above three components, and can be specifically calculated using the formula Y=0.299×R+0.587×G+0.114×B. The luminance image data is single-component image data composed of one type of component value (a value indicating luminance). The luminance component data includes luminance Y for each pixel based on the values (RGB values) of corresponding pixels in the scan data. Luminance component data is an example of first image data.

In S130, the CPU 210 performs edge extraction processing on the generated luminance image data to extract edges in the luminance image indicated by the luminance image data, thereby generating edge extraction data. Specifically, the CPU 210 applies a known edge extraction filter such as a Sobel filter to the value of each pixel of the luminance image data to calculate the edge strength of each pixel. The CPU 210 generates edge extraction data in which these edge intensities are values of a plurality of pixels.

In S140, the CPU 210 executes level correction processing on the edge extraction data to generate corrected edge extraction data. The level correction process is a correction process for enlarging a specific range within the range of gradation values (0 to 255 in this embodiment) that the pixel values of the edge extraction data can take.

FIG. 8 is a diagram showing an example of a tone curve for level correction processing. Specifically, the CPU 210 applies the tone curve of FIG. 8 to each pixel of the edge extraction data. As a result, all values equal to or higher than the threshold Vb (eg, 245) are converted to the maximum value (255), and all values equal to or lower than the threshold Va (eg, 10) are converted to the minimum value (0). Then, the range greater than the threshold Va and less than the threshold Vb is expanded to the range from 0 to 255. In this way, since the range including the binarization threshold (the range larger than the threshold Va and less than the threshold Vb in FIG. 8) is expanded before the binarization processing of S150, which will be described later, the binarization accuracy can be improved.

In S150, the CPU 210 executes binarization processing on the corrected edge extraction data to generate binary image data. For example, in the edge image data, the CPU 210 classifies pixels whose pixel values (that is, edge strength) are equal to or greater than a threshold (for example, 128) as edge pixels, and classifies pixels whose pixel values are less than the threshold as non-edge pixels. Classify as an edge pixel. In binary image data, as described above, edge pixels have a value of "1" and non-edge pixels have a value of "0".

According to the first binary image data generation process described above, as described with reference to FIG. 7, the scan image SI It is possible to reduce halftone dot features that appear in the image. Furthermore, as described with reference to FIG. 7, by executing edge enhancement processing on each of the plurality of pieces of smoothed component image data, the scan image SI smoothed by the smoothing processing edges can be properly emphasized. As a result, it is possible to appropriately identify edge pixels in the scan image SI while suppressing the identification of edge pixels caused by halftone dots.

Furthermore, since the luminance image data is used as the single-component image data, it is possible to more appropriately identify a plurality of edge pixels in the scan image SI. For example, halftone dots often have C, M, and Y primary colors used for printing. However, the brightness image data is relatively small. Therefore, by using the brightness image data, it is possible to appropriately suppress the edge pixels caused by halftone dots from being specified. In addition, there is often a relatively large difference in luminance between the color of the characters and the color of the background for easy reading of the characters. Therefore, by using luminance image data, it is possible to appropriately identify edge pixels indicating edges of objects such as characters.

Furthermore, in the edge enhancement process of FIG. 6, the mask value (also called a smooth value) corresponding to the target pixel is calculated (S210), and the difference ΔV between the value TV of the target pixel and the mask value corresponding to the target pixel is calculated ( S220), and calculation of the sum (TV+ΔV) of the value TV of the target pixel and the corresponding difference ΔV (S240). As a result, the edges of the scan image SI can be appropriately emphasized, so that omission of specifying edge pixels to be specified can be suppressed. As a result, edge pixels in the scanned image SI can be more appropriately identified.

Further, in the edge enhancement process of FIG. 6, among a plurality of pixels in the scanned image SI, pixels with a corresponding difference ΔV greater than or equal to the reference are subjected to unsharp mask processing, and the difference ΔV is less than the reference. are not subjected to unsharp mask processing (S230, 240). As a result, as described with reference to FIG. 7, it is possible to further suppress the enhancement of the difference in value between pixels caused by halftone dots in the scan image SI, so that it is possible to specify edge pixels caused by halftone dots. can be suppressed further. Edges of objects such as characters can be appropriately emphasized. Therefore, edge pixels in the scan image SI can be more appropriately identified.

A-5: Second Binary Image Data Generation Processing The second binary image data generation processing of S24 in FIG. 2 will be described. FIG. 9 is a flowchart of the second binary image data generation process. At S300, CPU 210 generates minimum component data using the scan data. Specifically, the CPU 210 obtains the minimum component value Vmin from each of the plurality of pixel values (RGB values) included in the scan data. The minimum component value Vmin is the minimum value among a plurality of component values (R value, G value, B value) included in the RGB values. The CPU 210 generates image data having these minimum component values Vmin as values of a plurality of pixels as minimum component data. The minimum component data is image data representing an image of the same size as the scanned image SI. Each of the plurality of pixel values included in the minimum component data is the minimum component value Vmin of the corresponding pixel values (RGB values) of the scan data.

FIG. 10 is an explanatory diagram of the minimum component value and maximum component value of scan data. FIGS. 10A to 10E show bar graphs of RGB values of cyan (C), magenta (M), yellow (Y), black (K), and white (W) as examples of RGB values. is illustrated. As shown in FIG. 10, the RGB values (R, G, B) of C, M, Y, K, and W are (0, 255, 255), (255, 0, 255), (255, 255, 0), (0, 0, 0), (255, 255, 255).

The luminance Y of these RGB values can be calculated using, for example, the formula Y=0.299*R+0.587*G+0.114*B, as described above. The luminances of C, M, Y, K, and W (represented by values from 0 to 255) are approximately 186, 113, 226, 0, and 255, which are different values (FIG. 10). On the other hand, the minimum component values Vmin of C, M, Y, K, and W are 0, 0, 0, 0, and 255, as shown in FIG. 10, which are the same except for white (W). .

FIG. 11 is a second diagram showing an example of an image used for image processing. FIG. 11A is an enlarged view of the above halftone dot area in the scanned image SI. For example, in the example of FIG. 11A, the halftone dot area in the scanned image SI includes a plurality of M dots MD and a plurality of Y dots YD. Here, for the sake of explanation, it is assumed that the image showing the M dots MD is a uniform image having the primary color of magenta, and the image showing the Y dots YD is a uniform image having the primary color of yellow.

FIG. 11B shows an example of the minimum component image MNI indicated by the minimum component data. This minimum component image MNI corresponds to the scanned image SI of FIG. 11(A). In the minimum component image MNI, the values of the pixels in the area MDb corresponding to the Y dots MD of the scan image SI are the same as the values of the pixels in the area YDb corresponding to the Y dots YD. FIG. 11C shows, as a comparative example, a luminance image YI represented by luminance image data indicating the luminance of each pixel. This luminance image YI corresponds to the scanned image SI in FIG. 11(A). In the luminance image YI, unlike the minimum component image MNI, the values of the pixels in the area MDd corresponding to the M dots MD of the scan image SI are different from the values of the pixels in the area YDd corresponding to the Y dots YD. .

As can be seen from the above explanation, in the minimum component image MNI, the difference between the values of the plurality of pixels corresponding to the portions in the document where the C, M, Y, and K dots are formed in the scanned image SI is It becomes smaller than the luminance image YI. In the minimum component image MNI, the values of the pixels in the ground color area corresponding to the area showing the ground color (the white of the paper) in the document in the scanned image SI are the values of the pixels corresponding to the portions where dots are formed. be larger than

In S310, the CPU 210 performs a smoothing process on the generated minimum component data to smooth the minimum component image MNI indicated by the minimum component data, thereby generating smoothed minimum component data. Specifically, the CPU 210 applies a predetermined smoothing filter, in this embodiment, a Gaussian filter of 5 pixels long by 5 pixels wide, to the value of each pixel of the minimum component data. Compute the value of a pixel. The smoothed minimum component data includes, for each pixel, a smoothed value generated by the above-described processing based on the corresponding pixel value (RGB value) of the scan data. The smoothed minimum component data is an example of second image data.

In S320, the CPU 210 executes an edge extraction process for extracting edges in the smoothed minimum component image MNI indicated by the smoothed minimum component data on the smoothed minimum component data, Generate edge extraction data. Specifically, the CPU 210 applies the same Sobel filter as in the process of S130 in FIG. 5 to the value of each pixel of the smoothed minimum component data to calculate the edge strength. The CPU 210 generates edge extraction data in which these edge intensities are values of a plurality of pixels.

In S330, the CPU 210 executes level correction processing on the edge extraction data to generate corrected edge extraction data. The level correction process is the same as the process of S140 in FIG. In S340, the CPU 210 executes the same binarization process as the process of S150 in FIG. 5 on the corrected edge extraction data to generate binary image data. In binary image data, as described above, edge pixels have a value of "1" and non-edge pixels have a value of "0".

According to the second binary image data generation processing described above, edge extraction processing is performed on the minimum component data to generate edge extraction data (S320). Then, a plurality of edge pixels of the scanned image SI are identified by executing edge pixel identification processing including binarizing the edge extraction data (S340) (S330, S340, S26 in FIG. 2). . With the minimum component data, as described with reference to FIG. 11, it is possible to suppress the difference in value between pixels in the halftone dot area. can be suppressed from being specified. Therefore, edge pixels in the scanned image SI can be appropriately identified.

More specifically, the elements that make up the halftone dot area are five types of C, M, Y, and K dots, and the ground color (white) of the paper. In the minimum component data, it is possible to suppress the difference in value between pixels representing four types of these elements. As a result, when using the minimum component data, it is possible to suppress the edge pixels indicating the edge of the halftone dot from being specified.

On the other hand, it is often the case that one of the character color and the background color has a dark color and the other has a light color. For this reason, if one of the characters and the background contains a relatively large portion indicating the ground color (white) of the paper, and the other contains a relatively large portion indicating the dots of C, M, Y, and K. There are many. As shown in FIG. 10, in the minimum component data, between the pixel values of the portion indicating the C, M, Y, and K dots and the pixel value of the portion indicating the background color (white) of the paper, There is a big difference. For this reason, if edge pixels are specified using the minimum component data, there is a high possibility that the edge pixels forming the edge of the character can be appropriately specified. In particular, yellow (Y) has a lower density (higher brightness) than C, M, and K. For this reason, when there is a yellow character on the background of the background color (white) of the paper, even if the brightness image data is binarized, the edge pixels forming the edge of the yellow character can be appropriately specified. Sometimes you can't. In this embodiment, even in such a case, it is possible to appropriately specify the edge pixels forming the edge of the yellow character. For this reason, in addition to identifying edge pixels using luminance image data, edge pixels are identified using minimum component data, thereby identifying edge pixels such as characters that cannot be identified using luminance image data alone. obtain. As a result, the accuracy of specifying edge pixels in the scanned image SI can be improved.

Furthermore, smoothing processing is performed on the minimum component data before edge extraction processing (S310). As a result, the smoothing process can further suppress the difference in value between pixels in the halftone dot area in the minimum component image MNI. For example, in the halftone dot area in the scanned image SI, the portion indicating the dot may not always be C, M, Y due to overlapping of C, M, Y, K dots, blurring during reading by the reading execution unit 290, or the like. , K primaries. For this reason, in the minimum component image MNI, the values between the plurality of pixels representing each dot of C, M, Y and K are small but not zero. A smoothing process can further reduce the difference in values between the pixels. As a result, the identification of edge pixels caused by halftone dots can be further suppressed. Also in the second binary image data generation process, as in the first binary image data generation process, the level correction process (S330) is executed, so that the accuracy of specifying the edge pixels in the scan image SI can be improved. can improve.

As described above, in the above embodiment, two types of single-component image data, that is, luminance image data and minimum component data, are used to finally identify edge pixels (S22 to S22 in FIG. 2). S26). Since a plurality of edge pixels in the scan image SI are identified using two types of single-component image data generated using different processing in this way, the plurality of edge pixels in the scan image SI can be It is possible to suppress specific omissions. For example, when there is a yellow character on a white background, the difference in brightness between white and yellow is relatively small, so it is difficult to specify the edge pixels forming the edge of the character using the brightness image data. is. On the other hand, as can be seen from FIGS. 10C and 10E, in the minimum component data, a large difference appears between white and yellow. Using the data, it is easy to identify the edge pixels that make up the edge of the character. Further, for example, when a yellow character is present on a magenta background, the minimum component data does not show a difference between magenta and yellow. is difficult to identify. On the other hand, when there is a yellow character on a magenta background, the difference in luminance between magenta and yellow is relatively large. It is easy to do.

Further, since the single-component image data used in addition to the luminance image data is the minimum component data, it is possible to suppress the edge pixels caused by halftone dots in the scanned image SI from being specified as described above.

A-6: Block Determination Processing The block determination processing of S28 in FIG. 2 will be described. 12 and 13 are flow charts of block determination processing of the first embodiment. FIG. 14 is an explanatory diagram of a plurality of blocks BL arranged on the scan image SI. As described above, the block determination process uses scan data to determine whether or not each of the plurality of blocks BL arranged in the scan image SI is a character block. This is the process to generate.

In S400, the CPU 210 prepares canvas data for generating block determination data in the memory (specifically, the buffer area of the volatile storage device 220). The canvas (initial image) indicated by the canvas data is an image having the same size as the scanned image SI, that is, an image having the same number of pixels. The value of each pixel of canvas data is a predetermined initial value (eg, 0).

In S402, the CPU 210 sets a plurality of blocks BL on the scanned image SI. For example, as shown in FIG. 14, the scan image SI is divided into blocks BL of P rows×Q columns arranged in a matrix (P and Q are integers of 2 or more). One block BL is a rectangular area containing m×n pixels (m and n are integers equal to or greater than 2). For example, m and n are each 20.

At S404, the CPU 210 selects one target block from a plurality of blocks BL set in the scan image SI.

In S405, the CPU 210 converts the RGB values of each pixel of the block of interest into color values of the HSV color system (also referred to as HSV values) using a known conversion formula. Each HSV value includes three component values: an H value that indicates hue, an S value that indicates saturation, and a V value that indicates brightness. The H value ranges from 0 degrees to 360 degrees. The S value ranges from 0% to 100%, with 0% indicating achromatic color. The V value ranges from 0% to 100%, with 0% indicating black and 100% indicating white.

In S407, the CPU 210 classifies each pixel in the block of interest into one of eight basic colors. The eight basic colors are K, C, M, Y, R, G, B, and W. For example, pixels whose S value is less than a predetermined reference value (eg, 20%) and whose V value is less than 50% are classified as K pixels. A pixel whose S value is less than a predetermined reference value and whose V value is 50% or more is classified as a W pixel. A pixel having an S value equal to or greater than a predetermined reference value is classified into one of C, M, Y, R, G, and B, which are chromatic colors, according to the H value (hue). The ranges of H values corresponding to C, M, Y, R, G, and B are predetermined.

In S410, the CPU 210 determines the most frequent color of the block of interest as the representative color (also called block color) of the block of interest. The most frequent color of the block of interest is the color for which the largest number of pixels are classified in S407 among the eight basic colors. In this embodiment, white (W), which is the color of the scanned paper, is assumed as the color of the background Bg1 of the scanned image SI. For this reason, the block color of the block BL that mainly includes the background Bg1 is determined to be W, and the representative color of the block BL that mainly includes an object different from the background Bg1 is C or M depending on the color of the object. , Y, K, R, G, or B.

In S415, the CPU 210 determines whether or not all blocks BL in the scan image SI have been processed as target blocks. If there is an unprocessed block BL (S415: NO), the CPU 210 returns the process to S404. If all blocks BL have been processed (S415: YES), the CPU 210 advances the process to S420.

In S420, CPU 210 generates color run length information RLc in the horizontal direction (first direction D1 in FIG. 14). The color run-length information RLc includes the number of consecutive blocks BL (also called background blocks) whose representative color is the background color (W in this embodiment) along the first direction D1, and the number of consecutive blocks BL. The number of consecutive blocks BL (also referred to as non-background blocks) whose representative colors are different colors (C, M, Y, K, R, G, and B in this actual example), and the recorded information. be. The color run-length information RLc is generated for each row of block groups (for example, block group BG1 in FIG. 14) extending from the left end to the right end of the scan image SI in FIG. In this embodiment, blocks BL of P rows×Q columns are arranged in the scanned image SI, so that one row of blocks BL includes Q blocks BL. Then, P pieces of color run length information RLc corresponding to the block group of P rows are generated.

FIG. 15 is an explanatory diagram of run length information. FIG. 15A shows an example of block group BGc for one row. The alphabet in each block BL indicates the representative color of that block BL. In FIG. 15A, hatched blocks are non-background blocks, and non-hatched blocks are background blocks. FIG. 15B shows an example of color run length information RLc. The color run length information RLc includes five consecutive number information Lc1 to Lc5. These consecutive number information Lc1 to Lc5 are, respectively, four consecutive background blocks A1, five consecutive non-background blocks A2, and two consecutive background blocks in the block group BGc in FIG. A3 corresponds to a group of three consecutive non-background blocks A4 and three consecutive background blocks A5.

At S430, the CPU 210 selects a non-background block of interest. Specifically, with reference to the color run-length information RLc, a set of non-background blocks continuous in the first direction D1 is selected from among the plurality of blocks BL arranged in the scan image SI as a non-background block of interest. is selected as For example, in the example of FIG. 14, the non-background blocks Aa and Ab corresponding to parts of the characters TX1 and TX2 in the scan image SI, the non-background block Ac corresponding to the horizontal line LN, and the characters BT larger than the characters TX1 and TX2 and a non-background block Ae corresponding to part of the photo PC. These non-background blocks are selected one by one as non-background blocks of interest.

In S435, the CPU 210 refers to the color run length information RLc to determine whether or not the number of consecutive non-background blocks of interest is less than the threshold TH1. In general, small characters are often composed of thin lines placed on the background, so a character area in which small characters are placed has an area with a background color and a non-background color (character color). region is repeated at short intervals in many cases. On the other hand, a halftone dot area in which a photograph or the like is arranged is a relatively large rectangular area, and the whole or most of the area has non-background colors (various colors in the photograph). For this reason, the number of consecutive non-background blocks corresponding to characters is smaller than the number of consecutive non-background blocks corresponding to halftone dot areas. The threshold TH1 is experimentally determined in advance using a sample image so that, for example, a character area in which small characters are arranged and a halftone dot area in which a photograph or the like is arranged can be discriminated.

If the number of consecutive non-background blocks of interest is less than the threshold TH1 (S435: YES), in S445 the CPU 210 determines the non-background block of interest to be a character block. If the number of consecutive non-background blocks of interest is equal to or greater than the threshold TH1 (S435: NO), in S450, the CPU 210 determines the non-background blocks of interest as halftone dot candidate blocks that are candidates for halftone dot blocks.

In S455, the CPU 210 determines whether or not all non-background blocks in the scan image SI have been processed as blocks of interest. If there is an unprocessed non-background block (S455: NO), the CPU 210 returns the process to S430. If all non-background blocks have been processed (S455: YES), the CPU 210 advances the process to S460.

At the time of S460, the plurality of blocks BL in the scan image SI are classified into any of background blocks, character blocks, and halftone dot candidate blocks. FIG. 16 is an explanatory diagram of various blocks specified on the scan image SI. In the example of FIG. 16, a plurality of blocks BA1 and BA2 corresponding to characters TX1 and TX2 in the scanned image SI are classified as character blocks. A plurality of blocks BA3 corresponding to horizontal lines LN, a plurality of blocks BA4 corresponding to large characters BT, and a plurality of blocks BA5 corresponding to photographs PC are classified as halftone dot candidate blocks. Blocks BL different from these blocks BA1 to BA5 are classified as background blocks.

Here, the appearance of the horizontal line LN is improved when the edges are emphasized in the same manner as the characters. Also, since the large character BT is a character, the appearance is improved when the edge is emphasized. For this reason, originally, blocks BL corresponding to horizontal lines LN and large characters BT should preferably be specified as character blocks, and should not be specified as halftone dot candidate blocks. At this stage, blocks BL corresponding to horizontal lines LN and large characters BT are identified as halftone dot candidate blocks, so it is preferable to correct the identification result so as to identify these blocks BL as character blocks. Therefore, in this embodiment, S460 to S515 in FIG. 13, which will be described below, are executed.

At S460, the CPU 210 labels the halftone dot candidate blocks to identify a halftone dot candidate area consisting of one or more consecutive halftone dot candidate blocks. Labeling is performed by assigning one identifier to one continuous region composed of one or more halftone dot candidate blocks continuous in the first direction D1 or the second direction D2, This is a process of assigning different identifiers to areas. In the example of FIG. 16, three halftone dot candidate areas are specified corresponding to the plurality of blocks BA3, the plurality of blocks BA4, and the plurality of blocks BA5.

At S465, the CPU 210 selects one halftone dot candidate area of interest from the one or more halftone dot candidate areas specified at S460.

At S470, the CPU 210 classifies each block BL forming the target halftone dot candidate area into either an edge block or a non-edge block. Specifically, the CPU 210 refers to the first binary image data generated in the first binary image data generation process (FIG. 5) of S22 of FIG. Count the number of 1 edge pixels. The CPU 210 classifies blocks BL whose number of first edge pixels is equal to or greater than a predetermined threshold value THe as edge blocks, and classifies blocks BL whose number of first edge pixels is less than a predetermined threshold value THe as non-edge blocks. do. In the modified example, referring to the second binary image data generated in the second binary image data generation process (FIG. 9) in S24 of FIG. 2, based on the number of second edge pixels, They may be classified into edge blocks and non-edge blocks. Both the first edge pixel and the second edge pixel are part of the character candidate pixels. Alternatively, the edge identification data generated by the logical sum synthesis processing of S26 in FIG. 2 may be referred to and classified into edge blocks and non-edge blocks based on the number of edge pixels (character candidate pixels). .

In FIG. 16, among a plurality of blocks BA3 to BA5 that are halftone dot candidate areas, cross-hatched blocks BL are edge blocks, and cross-hatched blocks BL are non-edge blocks. . For example, in the halftone dot candidate area (a plurality of blocks BA3) corresponding to the horizontal line LN, all blocks BL are classified as edge blocks. In the halftone dot candidate area (a plurality of blocks BA4) corresponding to the large character BT, only the blocks BL corresponding to the outer edge of the large character BT are classified as edge blocks. In the halftone dot candidate area (a plurality of blocks BA5) corresponding to the photograph PC, the block BL corresponding to the outer edge and some blocks BL inside are classified as edge blocks. The interior of the photo PC generally contains a mixture of smooth portions and edged portions. For this reason, in the halftone dot candidate area corresponding to the photo PC, edge blocks and non-edge blocks are scattered and mixed inside the photo PC.

At S475, the CPU 210 generates edge run length information RLe in the horizontal direction (the first direction D1 in FIG. 16) for the halftone dot candidate area of interest. The horizontal edge run-length information RLe is information in which the number of consecutive edge blocks (also referred to as the consecutive number) and the number of consecutive non-edge blocks along the first direction D1 are recorded. The horizontal edge run-length information RLe is a block group for one row extending from the left end to the right end (for example, the block generated for each group BG2). For example, if the halftone dot candidate area of interest includes a group of blocks of K rows (K is an integer equal to or greater than 1), then K pieces of horizontal edge run length information RLe corresponding to the group of blocks of K rows are generated. be.

FIG. 15C shows an example of a block group BGe for one line of the halftone dot candidate area. In the block group BGe, blocks BL marked with "E" are edge blocks, and blank blocks BL are non-edge blocks. FIG. 15D shows an example of the edge run length information RLe. The edge run length information RLe includes six consecutive number information Le1 to Lc6. These consecutive number information Le1 to Lc6 are respectively one edge block B1, three consecutive non-edge blocks B2, one edge block B3, and two consecutive non-edge blocks B3 in the block group BGe of FIG. 15(C). consecutive non-edge blocks B4, two consecutive edge blocks B5, and two consecutive non-edge blocks B6.

In S480, the CPU 210 generates edge run length information RLe in the vertical direction (the second direction D2 in FIG. 16) for the halftone dot candidate area of interest. The vertical edge run-length information RLe is information in which the number of consecutive edge blocks and the number of consecutive non-edge blocks are recorded along the second direction D2. The vertical edge run-length information RLe is a block group for one column extending from the top end to the bottom end of the halftone dot candidate area of interest (for example, the area consisting of a plurality of blocks BA5 in FIG. It is generated for each group BG3). For example, if the halftone dot candidate area of interest includes L columns of blocks (L is an integer equal to or greater than 1), L pieces of vertical edge run length information RLe corresponding to the L columns of blocks are generated. be.

In S485, the CPU 210 specifies the maximum consecutive number Vmx of non-edge blocks in the halftone dot candidate area of interest. Specifically, horizontal and vertical edge run-length information RLe is referred to, and the maximum number of horizontal and vertical continuations of non-edge blocks is specified as maximum continuation number Vmx.

In S490, the CPU 210 calculates the ratio Er of edge blocks in the halftone dot candidate area of interest. The ratio Er is a value obtained by dividing the number of edge blocks in the halftone dot candidate area of interest by the total number of blocks in the halftone dot candidate area of interest.

In S495, the CPU 210 determines whether or not the ratio Er occupied by edge blocks is greater than or equal to a predetermined threshold TH2. In halftone dot candidate regions corresponding to lines such as the horizontal line LN (FIG. 16) (for example, horizontal lines, vertical lines, and diagonal lines), most blocks are considered to be classified as edge blocks. For this reason, in halftone dot candidate areas corresponding to lines, the proportion Er occupied by edge blocks is higher than in halftone dot candidate areas corresponding to photographs and the like. The threshold TH2 is experimentally determined in advance using a sample image so that, for example, halftone dot candidate regions corresponding to lines and halftone dot candidate regions corresponding to photographs can be discriminated. The threshold TH2 is, for example, 80%.

If the ratio Er of edge blocks is greater than or equal to the predetermined threshold TH2 (S495: YES), in S505 the CPU 210 determines all blocks BL forming the target halftone dot candidate area as character blocks. As a result, blocks BL corresponding to lines (for example, a plurality of blocks BA3 in FIG. 16) are finally identified as character blocks. When the ratio Er occupied by edge blocks is less than the predetermined threshold TH2 (S495: NO), the CPU 210 advances the process to S500.

In S500, the CPU 210 determines whether or not the maximum consecutive number Vmx of non-edge blocks is greater than or equal to a predetermined threshold TH3. Large letters BT are generally solid objects. For this reason, edges exist on the outer edge of the large character BT, but few edges exist on the interior. In Photo PC, on the other hand, edges exist not only on the outer edge but also on the inside. Therefore, in halftone dot candidate areas corresponding to large characters BT, the maximum consecutive number Vmx of non-edge blocks is larger than in halftone dot candidate areas corresponding to photographs and the like. The threshold TH3 is experimentally determined in advance using a sample image so that, for example, a halftone dot candidate area corresponding to large characters BT and a halftone dot candidate area corresponding to a photograph can be discriminated.

If the maximum consecutive number Vmx of non-edge blocks is equal to or greater than the predetermined threshold TH3 (S500: YES), in S505 the CPU 210 determines all blocks BL forming the target halftone dot candidate area as character blocks. do. Blocks BL corresponding to large characters BT (for example, a plurality of blocks BA4 in FIG. 16) are finally identified as character blocks. If the maximum consecutive number Vmx of non-edge blocks is less than the predetermined threshold TH3 (S500: NO), in S510 the CPU 210 converts all blocks BL forming the target halftone dot candidate area into halftone dot blocks. decide. As a result, the blocks BL corresponding to the photograph PC (for example, the plurality of blocks BA5 in FIG. 16) are finally identified as halftone dot blocks.

In S515, the CPU 210 determines whether or not all halftone dot candidate areas in the scan image SI have been processed as the target halftone dot candidate area. If there is an unprocessed halftone dot candidate area (S515: NO), the CPU 210 returns the process to S465. If all halftone dot candidate areas have been processed (S515: YES), the CPU 210 advances the process to S520.

At the time of S520, all blocks BL in the scanned image SI are classified into character blocks, halftone dot blocks, or background blocks. In S520, the CPU 210 generates binary image data indicating character blocks and non-character blocks (dotted blocks or background blocks) as block determination data.

According to the present embodiment described above, the CPU 210 as the image acquisition unit acquires scan data generated by reading a printed matter as target image data (S10 in FIG. 2). The CPU 210 as a first specifying unit specifies a plurality of character candidate pixels out of a plurality of pixels in the scan image SI using the scan data (S22-S26 in FIG. 2). The CPU 210 as a continuation number identifying unit identifies the number of continuations of non-background blocks among the plurality of blocks BL arranged on the scan image SI (S420 in FIG. 12). The CPU 210 as a second identifying unit identifies character blocks from a plurality of blocks BL based on the number of continuations (S430 to S515 in FIGS. 12 and 13). The CPU 210 as a third identifying unit identifies a plurality of character pixels in the scan image SI based on the character candidate pixel identification result and the character block identification result (S29 in FIG. 2).

Objects that are different from the background have different feature continuity depending on the type, such as characters or photographs. For example, as described above, in an area containing characters, the background and thin lines appear alternately, so the number of consecutive non-background blocks tends to be smaller than in the area of the photo PC. According to this embodiment, character blocks are identified based on the number of consecutive non-background blocks, so character blocks can be appropriately identified in consideration of the continuity of the features described above. Therefore, the character pixels in the scan image SI can be accurately identified based on the character candidate pixel identification results and the character block identification results. For example, it is possible to reduce the inconvenience of erroneously identifying edges of objects and edges of halftone dots in the photo PC as character pixels.

Furthermore, according to the present embodiment, the CPU 210 as the third specifying unit specifies, among the plurality of specified character candidate pixels, pixels located in the specified character block as character pixels (see FIG. 2 S29). As a result, character pixels can be specified with high accuracy from among a plurality of character candidate pixels.

Furthermore, according to the present embodiment, as the characteristic value of each block BL, values relating to the colors of a plurality of blocks, specifically, representative colors of each block BL (eight values (C, M, Y, K, R, G, B, W)) is used. A non-background block is a block having a color different from the background color (W) of the scan image SI. An object having a color different from the background color has different color continuity depending on the type, such as a character or a photograph. For example, in an area containing characters, areas of a color different from the background color (character color) and areas of the background color alternately appear corresponding to the alternating appearance of the background and thin lines. On the other hand, in the Photo PC area, areas with a color different from the background color tend to be continuous. For this reason, the number of consecutive non-background blocks having a color different from the background color is likely to be smaller in the area containing the characters than in the area of the photo PC. According to this embodiment, character blocks can be specified with high accuracy by utilizing the difference in color characteristics between characters and photographs.

Furthermore, according to the present embodiment, the CPU 210 as the second specifying unit determines (see FIG. 12, YES at S435), the pair of non-background blocks are identified as character blocks (S445 in FIG. 12). As described above, since characters are composed of thin lines arranged on the background, the number of consecutive non-background blocks corresponding to characters (corresponding to lines forming characters) is likely to be relatively small. In this embodiment, such characteristics can be used to specify character blocks with high accuracy.

According to this embodiment, when the number of continuous non-background blocks in a set of continuous non-background blocks is equal to or greater than the threshold TH1 (NO in S435 of FIG. 12), the CPU 210 as the second specifying unit For each non-background block (block BL forming a halftone dot candidate area), an index value (for example, maximum consecutive number Vmx or ratio Er) for character candidate pixels (for example, first edge pixels) included in the non-background block is calculated, and if the index value satisfies a specific condition, the set of non-background blocks is specified as the character blocks (S460 to S515 in FIG. 13).

Even if the consecutive number of a set of non-background blocks is equal to or greater than the threshold TH1, depending on the distribution of character candidate pixels included in the non-background blocks, the object corresponding to these blocks BL may not be a photograph PC but a large It may be a character BT or a line (for example, a horizontal line LN) to which image processing similar to that of the character should be applied. According to the present embodiment, even in such a case, the character block can be specified with high accuracy using the index values for the character candidate pixels included in the non-background block.

More specifically, the CPU 210 as the second specifying unit determines the threshold value for each of a set of non-background blocks whose number of continuations is equal to or greater than the threshold value TH1, that is, a plurality of blocks BL forming the halftone dot candidate region. It is determined whether the edge block includes THe (also referred to as a reference number) or more of the first edge pixels (which are part of the character candidate pixels) (S470 in FIG. 13), and a set of non-background A ratio Er of edge blocks to blocks is calculated as an index value (S490 in FIG. 13). When the ratio Er of edge blocks is equal to or greater than a predetermined threshold TH2 (also referred to as a reference ratio) (YES in S495 of FIG. 13), the CPU 210 identifies a set of non-background blocks as character blocks ( S505). As mentioned above, non-background blocks corresponding to lines have a relatively high proportion of edge blocks corresponding to edges of the line. In this embodiment, by using such characteristics, it is possible to accurately identify non-background blocks corresponding to lines as character blocks.

Furthermore, the CPU 210 as a second specifying unit determines the number of consecutive non-background blocks equal to or greater than the threshold TH1, that is, the plurality of blocks BL constituting the halftone dot candidate region, each of which is less than the threshold THe. 13 (S470 in FIG. 13). The maximum consecutive number Vmx of consecutive edge blocks in the horizontal direction or the vertical direction is calculated as an index value (S485 in FIG. 13). CPU 210 identifies a set of non-background blocks as character blocks when the maximum consecutive number Vmx is equal to or greater than a predetermined threshold TH3 (also referred to as a third criterion) (YES at 500 in FIG. 13). S505). As described above, non-background blocks corresponding to large characters BT have a larger maximum consecutive number Vmx than non-background blocks corresponding to photographs PC and the like. In this embodiment, using such characteristics, a set of non-background blocks corresponding to the large character BT can be specified as character blocks with high accuracy.

B. Second Embodiment In the second embodiment, the contents of the block determination processing in S28 of FIG. 2 are different from those in the first embodiment. Other processing and configuration of the second embodiment are the same as those of the first embodiment. FIG. 17 is a flow chart of block determination processing of the second embodiment. In the block determination process of FIG. 17, the processes of S460 to S515 of FIG. 13 are not executed. Instead, in the block determination process of FIG. 17, the process of S440B is added after S435.

The processing of S400 to S420 of the block determination processing of FIG. 17 is the same as the processing of S400 to S420 of FIG. In S430, the CPU 210 selects a non-background block of interest as in S430 of FIG. In S435, as in S435 of FIG. 12, the CPU 210 refers to the color run length information RLc to determine whether or not the number of consecutive non-background blocks of interest is less than the threshold TH1.

If the number of consecutive non-background blocks of interest is less than the threshold TH1 (S435: YES), in S445, the CPU 210 determines the non-background block of interest to be a character block, as in S445 of FIG. If the number of consecutive non-background blocks of interest is equal to or greater than the threshold TH1 (S435: NO), in S440B the CPU 210 determines whether the number of colors of the non-background blocks of interest is less than a predetermined threshold THc. . The non-background block of interest is composed of blocks BL whose number is equal to or greater than the threshold TH1, and the representative color of each of these blocks is determined to be one of C, M, Y, K, R, G, and B. (S410). The number of representative colors included in the number of blocks BL constituting the non-background block of interest that is equal to or greater than the threshold TH1 is specified as the number of colors. It is determined whether or not the number of colors is less than a predetermined threshold THc.

In general, lines such as large letters BT and horizontal lines LN are often solidly painted in one color. Photo PCs, on the other hand, often contain a variety of colors. For this reason, when the non-background block of interest corresponds to a large character BT or a line such as a horizontal line LN, the number of colors of the non-background color block of interest is smaller than when the non-background block of interest corresponds to the photograph PC. . The threshold THc is experimentally determined in advance using a sample image so that, for example, non-background blocks corresponding to large characters and lines can be distinguished from non-background blocks in which photographs and the like are arranged. The threshold THc is 1, for example.

When the number of colors of the target non-background block is less than the predetermined threshold value THc (S440B: YES), in S445, the CPU 210 determines all blocks BL forming the target non-background block as character blocks. Blocks BL corresponding to lines and blocks BL corresponding to large characters are thereby identified as character blocks. If the number of colors of the target non-background block is equal to or greater than the predetermined threshold value THc (S440B: NO), in S450B the CPU 210 determines all blocks BL forming the target non-background block as halftone dot blocks. . As a result, the block BL corresponding to the photo PC is identified as a halftone dot block.

In S455, the CPU 210 determines whether or not all non-background blocks in the scan image SI have been processed as blocks of interest. If there is an unprocessed non-background block (S455: NO), the CPU 210 returns the process to S430. If all non-background blocks have been processed (S455: YES), the CPU 210 advances the process to S520.

At the time of S520, the plurality of blocks BL in the scan image SI are classified into any of background blocks, character blocks, and halftone dot blocks. In S520, similarly to S520 in FIG. 13, the CPU 210 generates binary image data indicating character blocks and non-character blocks (halftone blocks or background blocks) as block determination data.

According to the second embodiment described above, the CPU 210 as the second specifying unit determines that the continuous number of a set of continuous non-background blocks (non-background blocks of interest) is equal to or greater than the threshold TH1 (also referred to as the first criterion). (NO in S435 of FIG. 16) and the number of colors of a set of non-background blocks is less than a threshold THc (also referred to as a second criterion) (YES in S440B of FIG. 16), A set of non-background blocks are identified as character blocks (S445 in FIG. 16).

Even if the number of consecutive non-background blocks is equal to or greater than the threshold TH1, if the number of colors is small, the object in these blocks BL is not a photograph but is subjected to large characters BT or image processing similar to characters. It is likely to be a power line (eg, horizontal line LN). According to the present embodiment, it is possible to accurately identify a non-background block corresponding to a large character BT or a line as a character block by using such a characteristic of the number of colors.

C. Variant:
(1) In the above embodiment, a value indicating a representative color (one of the 8 values of CMYKRGBW) is adopted as the feature value of the block BL, and the block BL whose representative color is white (W) is specified as a background block, A block BL whose representative color is a color other than white is specified as a non-background block (S404 to S420 in FIG. 12). Not limited to this, for example, when a document made of paper other than white is used, a block BL having a characteristic value (e.g., average color value) indicating the color of the paper is specified as a background block, and the color of the paper is specified. A block BL having a feature value indicating a color different from that may be identified as a non-background block.

Further, as the characteristic value of the block BL, the average lightness, the average luminance, the variation of the pixel values, etc. may be employed. For example, a block BL whose average brightness or average brightness is equal to or higher than a standard may be specified as a background block, and a block BL whose average brightness or average brightness is less than a standard may be specified as a non-background block. Alternatively, a block BL whose pixel value variation is less than a reference may be identified as a background block, and a block BL whose pixel value variation is greater than or equal to the reference may be identified as a non-background block.

(2) In the above embodiment, among the plurality of character candidate pixels identified in S26 of FIG. 2, pixels located in the character block identified in S28 of FIG. 2 are identified as character pixels. (S29 in FIG. 2). The character pixels may be finally identified by other methods based on the character candidate pixel identification result and the character block identification result. For example, a character candidate pixel included in a character block is identified as a character pixel if it has a relatively wide range of colors that can be characters (e.g., black, red, blue, yellow, etc.), A character candidate pixel contained in a non-character block may be identified as a character pixel only if it has a color that is likely to be a character (eg, only black).

(3) In the above embodiment, when the number of consecutive non-background blocks in a set is less than the threshold TH1, the set of non-background blocks is specified as a character block. Alternatively, for example, if the size of the block BL to be set is significantly smaller than the lower limit of the character size to be specified, the number of consecutive non-background blocks set may exceed the predetermined lower limit. The set of non-background blocks may be specified as character blocks only when they are equal to or greater than the value and less than a predetermined upper limit.

(4) In the first binary image data generation process (FIG. 5), luminance image data is used as single-component image data (S120). Instead of this, average component value image data in which the average value of three component values (R value, G value, B value) included in the RGB values of the corresponding pixel of the scan data is used as the value of each pixel is used. It's okay to be

(5) In the second binary image data generation process (FIG. 9) of the above embodiment, minimum component data is used as single component image data (S300). Alternatively, maximum component data or inverted minimum component data may be used.

The maximum component data includes a plurality of values corresponding to a plurality of pixels included in the scan data, each of the plurality of values being the maximum component value Vmax of the corresponding pixels of the scan data. The maximum component value Vmax is the maximum value among a plurality of component values (R value, G value, B value) included in the RGB values of the corresponding pixel of the scan data.

Inverted minimum component data is obtained as follows. First, an inverted color value is generated by inverting a plurality of component values (R value, G value, B value) for each of a plurality of pixel values (RGB values) included in scan data. Assuming that the RGB values before inversion are (Rin, Gin, Bin), the inverted RGB values (Rout, Gout, Bout) are represented by the following equations (1) to (3).

Rout=Rmax-Rin (1)
Gout=Gmax-Gin (2)
Bout=Bmax-Bin (3)

Here, Rmax, Gmax, and Bmax are the maximum possible values of the R value, G value, and B value, respectively, and Rmax=Gmax=Bmax=255 in this embodiment. Image data having these inverted RGB values as the values of a plurality of pixels is generated as inverted image data. Then, using the inverted image data, inverted minimum component data is generated. Specifically, the inverted minimum component value VRmin is acquired from each of the plurality of inverted RGB values included in the inverted image data. The inverted minimum component value VRmin is the minimum value among a plurality of component values (R value, G value, B value) included in the inverted RGB values. The inverted minimum component data is image data having these inverted minimum component values VRmin as values of a plurality of pixels.

The inverted minimum component value VRmin is an inverted value of the maximum component value, and the relationship VRmin=(255-Vmax) is established. For this reason, both the maximum component data and the inverted minimum component data are values based on the maximum value among a plurality of component values included in the value of each pixel of the scan data (the inverted value of the maximum value, or The maximum value itself) can be said to be image data having a pixel value.

As shown in FIG. 10, the maximum component values Vmax of C, M, Y, K, and W are 255, 255, 255, 0, and 255, which are the same except for black (K). Therefore, in the maximum component data and the inverted minimum component data, one of the five types of elements forming the halftone dot area, that is, each dot of C, M, Y, and K, and the background color of the paper (white) Differences in values between pixels representing the four types of elements (C, M, Y dots and paper ground color (white)) are suppressed. As a result, when using the maximum component data or the inverted minimum component data, it is possible to suppress the edge pixels caused by halftone dots from being identified, as in the case of using the minimum component data.

(6) In the image processing of FIG. 2 of the above embodiment, the second binary image data generation processing of S24 and the logical sum synthesis processing of S26 may be omitted. That is, the first binary image data generated in the first binary image data generation process of S22 may be the final edge identification data. Conversely, the first binary image data generation process of S22 and the logical sum synthesis process of S26 may be omitted. That is, the second binary image data generated in the second binary image data generation process of S24 may be the final edge identification data.

(7) In the above embodiment, character sharpening processing is executed for character pixels (S40 in FIG. 2), and halftone dot smoothing processing is executed for non-character pixels (S30 in FIG. 2). . Alternatively, text pixels may be subjected to anti-aliasing to improve the appearance of the text. For non-character pixels, for example, a process of skipping colors (a process of converting to white) may be executed in order to reduce the amount of color material used during printing. In general, it is preferable to perform different image processing on character pixels and non-character pixels. Alternatively, specific image processing may be performed on either character pixels or non-character pixels, and the specific image processing may not be performed on the other.

(8) In the above embodiment, a Sobel filter is used in the edge extraction processing of S130 of FIG. 5 and S320 of FIG. Alternatively, other edge extraction filters such as a Roberts filter or a Laplacian filter may be used in these edge extraction processes.

(9) In the above embodiments, the target image data is scan data, but is not limited to this. The target image data may be generated by reading printed matter with a digital camera equipped with a two-dimensional image sensor.

(10) In the above embodiment, the edge identification data is generated by ORing the first binary image data and the second binary image data (S26 in FIG. 2). Alternatively, the edge identification data may be generated by ORing the first binary image data, the second binary image data, and the third binary image data. For the third binary image data, for example, binary image data generated using the maximum component data described above may be used. As a result, it is possible to further suppress omissions in identifying edges such as characters.

(11) The first binary image data generation process (FIG. 5) and the second binary image data generation process (FIG. 9) of the above embodiments can be changed as appropriate. For example, all or part of the processing of S110 and S140 in FIG. 5 can be omitted. Also, all or part of S310 and S330 in FIG. 9 can be omitted.

(12) The image processing apparatus that implements the image processing in FIG. 2 is not limited to the multifunction machine 200, and may be various apparatuses. For example, the image processing of FIG. 2 may be executed by a scanner or digital camera to generate print data to be supplied to a printer using image data generated by itself. Further, for example, a terminal device (for example, the terminal device 100) or a server (not shown) connected to be able to communicate with a scanner or printer executes the image processing of FIG. 2 using the scan data acquired from the scanner. may generate print data and supply the print data to the printer. Also, a plurality of computers (for example, a cloud server) that can communicate with each other via a network may share functions required for image processing, and execute image processing as a whole. In this case, the entirety of a plurality of computers is an example of an image processing device.

(13) In each of the above embodiments, part of the configuration implemented by hardware may be replaced with software, or conversely, part or all of the configuration implemented by software may be replaced with hardware. You may do so. For example, the logical sum synthesis processing of S26 and the logical product synthesis processing of S29 of FIG. 2 may be executed by dedicated hardware such as ASIC.

Although the present invention has been described above based on examples and modifications, the above-described embodiments of the present invention are intended to facilitate understanding of the present invention, and are not intended to limit the present invention. The present invention may be modified and modified without departing from the spirit and scope of the claims, and the present invention includes equivalents thereof.

DESCRIPTION OF SYMBOLS 100... Terminal device, 200... MFP, 210... CPU, 220... Volatile memory device, 230... Non-volatile memory device, 240... Display part, 250... Operation part, S270... Communication IF, 280... Print execution part, 290 ... reading execution part, PG ... computer program, SI ... scanned image, TI ... character identification image, GI ... smoothed image, FI ... processed image, EI ... edge identification image, BI ... block determination image, YI ... brightness image, BL...Block

Claims (12)

An image processing device,
an image acquisition unit that acquires target image data generated by reading printed matter using an image sensor;
A first identifying unit that uses the target image data to specify a plurality of candidate pixels that are candidates for character pixels from among a plurality of pixels in the target image indicated by the target image data , the first identification unit including a process of identifying a plurality of edge pixels forming an edge in the target image as the candidate pixels;
An arrangement unit for arranging a plurality of blocks in the target image, wherein the plurality of blocks are different from a plurality of background blocks having specific feature values indicating characteristics of the background and the specific feature values. a non-background block having feature values;
a continuation number identifying unit that identifies a continuation number, which is the number of continuations of non-background blocks among a plurality of blocks arranged in the target image, using the target image data;
a second identifying unit that identifies a character block from the plurality of blocks based on the number of continuations;
a third identifying unit that identifies the plurality of character pixels in the target image based on the result of identifying the candidate pixel and the result of identifying the character block;
An image processing device comprising: The image processing device according to claim 1,
The image processing device, wherein the third specifying unit specifies, as the character pixel, a pixel located in the specified character block among the plurality of specified candidate pixels. The image processing device according to claim 1 or 2,
The specific feature value is a value indicating the color of the background ,
The image processing device, wherein the non-background block is a block having a feature value indicating a color different from the color of the background . The image processing device according to any one of claims 1 to 3,
the second identifying unit identifies the set of consecutive non-background blocks as the character block when the number of consecutive sets of the non-background blocks is less than a first reference; Image processing device. The image processing device according to any one of claims 1 to 4,
The second specifying unit
the number of consecutive non-background blocks is equal to or greater than a first criterion and the number of colors of the consecutive non-background blocks is less than a second criterion; The image processing device that specifies the set of non-background blocks to be the character blocks as the character blocks. The image processing device according to any one of claims 1 to 4,
The second specifying unit
an index relating to the candidate pixels included in the non-background blocks for each of the set of consecutive non-background blocks when the number of consecutive ones of the set of consecutive non-background blocks is equal to or greater than a first criterion; Calculate the value of
An image processing device that specifies the set of continuous non-background blocks as the character blocks when the index value satisfies a specific condition. The image processing device according to claim 6,
The second specifying unit
determining whether each of the continuous set of non-background blocks is a candidate block containing a reference number or more of the candidate pixels;
calculating, as the index value, the ratio of the candidate blocks in the set of consecutive non-background blocks ;
The image processing device, wherein when the ratio of the candidate blocks is equal to or greater than a reference ratio, the set of continuous non-background blocks is specified as the character blocks. The image processing device according to claim 6,
The second specifying unit
determining whether each of the continuous set of non-background blocks is a candidate block containing the candidate pixels equal to or greater than a reference number or a non-candidate block containing the candidate pixels less than the reference number;
calculating, as the index value, a maximum consecutive number of consecutive non-candidate blocks in a predetermined direction in the set of consecutive non-background blocks ;
The image processing device, wherein the set of consecutive non-background blocks is specified as the character block when the maximum consecutive number is equal to or greater than a third criterion. The image processing device according to any one of claims 1 to 8,
The first specifying unit
using the target image data to generate a plurality of first values corresponding to the values of the plurality of pixels included in the target image data, based on the values of the corresponding pixels of the target image data; generating first image data including the plurality of first values calculated by the first process ;
identifying a plurality of first edge pixels indicative of edges in a first image represented by the first image data;
using the target image data to generate a plurality of second values corresponding to the values of the plurality of pixels included in the target image data, based on the values of the corresponding pixels of the target image data; generating second image data including the plurality of second values calculated by a second process different from the first process ;
identifying a plurality of second edge pixels indicative of edges in a second image represented by the second image data;
including a plurality of pixels corresponding to the plurality of first edge pixels and a plurality of pixels corresponding to the plurality of second edge pixels among the plurality of pixels in the target image; A group of pixels that does not include a plurality of pixels corresponding to neither the plurality of first edge pixels nor the plurality of second edge pixels is specified as the plurality of candidate pixels. , image processor. The image processing device according to any one of claims 1 to 9, further comprising:
performing a first image process on the values of the specified character pixels in the target image data, and performing a second image process different from the first image process on the values of the pixels different from the character pixels; and an image processing unit configured to generate the image-processed target image data. The image processing device according to claim 10, further comprising:
An image processing apparatus comprising a print data generation unit that generates print data using the image-processed target image data. A computer program,
an image acquisition function for acquiring target image data generated by reading printed matter using an image sensor;
A first specifying function of specifying a plurality of candidate pixels, which are candidates for character pixels, from among a plurality of pixels in a target image indicated by the target image data, using the target image data, the first specifying function including a process of specifying a plurality of edge pixels forming an edge in the target image as the candidate pixels;
An arrangement function for arranging a plurality of blocks in the target image, wherein the plurality of blocks are a plurality of background blocks having a specific feature value that indicates a feature of the background, and a plurality of background blocks that differ from the specific feature value. a non-background block having feature values;
a continuation number identification function that identifies a continuation number, which is the number of continuations of non-background blocks among a plurality of blocks arranged in the target image, using the target image data;
a second specifying function for specifying a character block from the plurality of blocks based on the number of continuations;
a third specifying function of specifying a plurality of the character pixels in the target image based on the result of specifying the candidate pixel and the result of specifying the character block;
A computer program that makes a computer realize

Images