JP5206468B2 - Image processing apparatus and image processing program - Google Patents

Description

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

  The received image is divided into, for example, a pattern region and a character region, and different image processing is performed on each region.

  As a technology related to this, for example, Patent Document 1 has a problem of increasing the compression ratio of digital image data, and the original pixel map is segmented into two planes, and then each plane is separated in an efficient manner. The pixels that are compressed to lossy compression are placed in one plane, the pixels that are to be losslessly compressed are placed in the other plane, and the selector plane is all pixels in the original pixel map. It is disclosed that it is created for tracking purposes.

  Further, for example, Patent Document 2 has an object to provide a system and method for converting a document into image data in a composite raster content format having a plurality of binary foreground planes. The image data is analyzed line by line. Identify pixels that have similar color and edge features, and those pixels on one line and pixels on another line are linked if they share similar features, and these linked It is disclosed that pixels are grouped as a region, an average color of this region is calculated, and an image map specifying the region to which the pixel belongs is generated along with the final table of regions.

  Also, for example, Patent Document 3 discloses that image separation processing, particularly character line drawing separation, is performed satisfactorily, image quality deterioration is reduced when resolution conversion processing and compression processing are performed, and image quality is high and compression ratio is high. An image processing apparatus and an image processing method capable of transmitting or storing image data. When image data is input, the image data is converted into a predetermined color space by a color space conversion unit and then separated by a selection data generation unit. Character line drawing information is extracted from the plane, and the expansion processing unit performs expansion processing for thickening the character line drawing on the separation information plane, generates an expansion separation information plane, and uses this expansion separation information plane to create a multi-layer separation unit. The color information of the character / line drawing part is separated into the character information plane, the pattern information other than the character / line drawing part and the background image are separated into the pattern information plane, and separated using the expansion separation information plane. Guo portion is disclosed to be prevented from being separated into picture information plane.

  Further, for example, in Patent Document 4, in the image area separation, with respect to characters or the like written on a color area such as light blue that can extract halftone dots, the entire area is extracted as halftone dots. The character cannot be extracted as an image area separation result. First, foreground and background are separated from the input image data by the binarization unit, and the predetermined number of pixels is obtained by the line segment extraction unit. The process of extracting the line segment of the minute width is performed, the same label is given to the connected pixels that are set as the foreground in the binarization unit, and for each label area to which the same label is given, The label feature value calculation unit calculates an image feature value, the label attribute value for each label region is determined by the first attribute determination unit, and the label feature value for each label region labeled by the labeling unit is determined. Attribute determination by one attribute determination unit With fruit, to perform the correction of the line segment extracted by the line segment extraction section in the segment corrector is disclosed.

  Further, for example, in Patent Document 5, when there are a wide part and a thin part in one character in image area separation, there is a part that is not an extracted part, and there is an error in the pattern. If extraction occurs, block-like erroneous extraction may occur, and image quality degradation is likely to be noticeable when the image is reproduced. First, a region with high density is input to the input image data by the shadow extraction unit. And a line segment having a width corresponding to the first predetermined number of pixels is extracted by the line segment extraction unit, and the shadow extraction unit and the network are extracted. A region having a width corresponding to the second predetermined number of pixels is extracted from each region extracted by the point extraction unit by the small region extraction unit, and a line segment extraction unit is extracted by the line segment correction unit based on the extracted region. Line extracted by It is disclosed that perform the correction.

  Also, for example, in Patent Document 6, encoding is performed satisfactorily regardless of whether the image to be encoded is a photographic image or a character image. For the purpose of providing an image encoding device that can perform encoding according to the characteristics of an image even when the edge of the block is small or when a slight edge is applied to the block boundary, 1 block extraction unit extracts a pixel block, the region determination unit for the extracted pixel block based on features such as dispersion information, histogram information, dynamic range information, whether the pixel block is photographic, If the pixel block is photographic, the pixel block is transform-coded in the transform coding unit. If the pixel block is character, the block is It is disclosed that is the block truncation coding by truncation coding section.

  In addition, for example, Patent Document 7 aims to provide an image area identification method that can identify each area of a mixed image of color characters / halftones with high performance and has excellent hardware feasibility. A pixel blocking unit; a characteristic value detecting unit that detects a plurality of characteristic values from pixels in the block; and a region identifying unit that performs character / halftone region identification from the plurality of characteristic values. The average value in the block and the total value in each block of the high-level pixel and the intermediate-level pixel are three kinds, and the area identification means includes a plurality of characteristic value quantization means and an intermediate level for area identification in advance. A LUT in which a threshold value of the sum total value of pixels is set, and a comparison unit, and a threshold value for the sum value of intermediate level pixels from the LUT using a value obtained by quantizing and combining the average value in the block and the sum value of the high level pixels as an address Look, to identify the region of to the block comparison actually obtained and the total value of the intermediate-level pixels is disclosed.

  Further, for example, Patent Document 8 has an object of satisfactorily separating an image into two types of regions based on the local characteristics, and an image signal input from the image input unit is stored in the image storage unit. The feature discriminating unit reads the image from the image storage unit in units of pixels or for each rectangular area of a predetermined size, performs binary discrimination for the predetermined feature, stores the result in the discrimination result storage unit, and The determination unit determines region information based on the determination result stored in the determination result storage unit, and the output selection unit sequentially reads out pixel data in the document storage unit, and 2 according to the region information determined by the region determination unit. It is disclosed that two areas of image output including an image signal including a first area and an image signal including a second area are obtained by distributing the two areas.

JP 2000-196893 A JP 2005-228331 A JP 2000-194857 A JP 2001-127999 A JP 2001-144554 A Japanese Patent Laid-Open No. 06-319133 JP 04-317261 A JP 2000-261663 A

  In the present invention, when determining the attribute of an area obtained by dividing an image, an erroneous determination that an area belonging to the attribute does not belong, or an erroneous determination that an area not belonging to the attribute belongs An object of the present invention is to provide an image processing apparatus and an image processing program that suppress determination.

The gist of the present invention for achieving the object lies in the inventions of the following items.
The invention according to claim 1 includes a dividing unit that divides the received image into a plurality of regions in an arbitrary shape, and a first determining unit that determines whether the region divided by the dividing unit belongs to a certain attribute. A second determination means for determining whether or not the area divided by the dividing means belongs to the attribute by a determination method different from the determination method performed by the first determination means; and the first determination based on the determination result by the result and the second determination means determines by means said areas divided by the dividing means comprises a third determining means for determining whether belonging to the attribute, the third The determination means determines that the second area determined by the first determination means as not belonging to the attribute but determined as belonging to the attribute by the second determination means is the first determination means by the first determination means. attribute An image processing apparatus and determines that belongs to the second region in the attribute when in contact even partially in the first region is a region which is determined to belong.

According to a second aspect of the present invention, the first determination unit and the second determination unit are a pixel block that is a pixel group in which pixels having the same pixel value, a pixel value frequency, and a pixel value in the region are connected. The image processing apparatus according to claim 1, wherein a determination is made as to whether or not the area belongs to the attribute based on any one or more of the numbers.

According to a third aspect of the present invention, the first determining means determines whether or not a region belongs to the attribute by comparing a maximum value of a histogram of pixel values with the first threshold value, and the second threshold value . The determination means uses a second threshold value different from the first threshold value used by the first determination means to determine whether a region belongs to the attribute as a maximum value of a histogram of pixel values. the image processing apparatus according to any one of claims 1 or 2, characterized in that by comparison with the second threshold value.

According to a fourth aspect of the present invention, pixels having the same pixel value are connected based on a ratio of pixels included in a region belonging to the attribute and pixels included in a region not belonging to the attribute. the image processing apparatus according to any one of claims 1 to 3, characterized in that pixel block is a pixel group are having a pixel block judgment means for judging whether the attribute.

The invention of claim 5, wherein the pixel block judging means, if is included entirety of the pixel block in the region belonging to the attributes, characterized by determining that該画pixel block is the attribute An image processing apparatus according to claim 4 .

The invention of claim 6, wherein the pixel block determining means, if it contains a portion of the pixel block in the region belonging to the attributes,該画pixel block is a feature that determined not to be the attribute The image processing apparatus according to claim 4 .

According to a seventh aspect of the present invention, there is provided a dividing unit that divides the received image into a plurality of regions in an arbitrary shape, and a computer that determines whether the region divided by the dividing unit belongs to a certain attribute. A determination means, a second determination means for determining whether or not the area divided by the dividing means belongs to the attribute by a determination method different from the determination method performed by the first determination means, based on the determination result by the result and the second determination means determines according to the first determining means, areas divided by said dividing means to function as a third determination means for determining whether or not belonging to the attribute, the The third determination means determines that the second area determined not to belong to the attribute by the first determination means but determined to belong to the attribute by the second determination means is the first determination. An image processing program, characterized in that determined to belong to regions of the second to the attribute when in contact even partially in the first region is a region which is judged as belonging to the attributes by means.

According to the image processing apparatus of claim 1, when determining an attribute of an area obtained by dividing an image, whether the attribute belongs to an attribute using only one of the first determination unit and the second determination unit Compared with the case of determining whether or not, it is possible to suppress an erroneous determination that an area belonging to the attribute does not belong, or an erroneous determination that an area that does not belong to the attribute belongs. In determining the attribute of a region having an area larger than the region obtained by dividing the image, it is determined whether only one of the first determination unit and the second determination unit belongs to a certain attribute. Compared to the above, errors in the determination can be suppressed.

According to the image processing apparatus of the second aspect , it is possible to make an efficient determination as compared with the case where the present configuration is not provided.

According to the image processing apparatus of the third aspect , it is possible to reduce the size and speed of the apparatus as compared with the case where this configuration is not provided.

According to the image processing apparatus of the fourth aspect , it is possible to make a more precise determination than when the present configuration is not provided.

According to the image processing apparatus of the fifth aspect , it is possible to increase the possibility that it is determined that the attribute is not a certain attribute as compared with the case where the present configuration is not provided.

According to the image processing apparatus of the sixth aspect , it is possible to increase the possibility of determining that the attribute is a certain attribute as compared with the case where the present configuration is not provided.

According to the image processing program of claim 7 , when determining the attribute of the area into which the image is divided, it is erroneous that the area belonging to the attribute does not belong as compared with the case where this configuration is not provided. It is possible to suppress a determination or an erroneous determination that an area that does not belong to the attribute belongs. In determining the attribute of a region having an area larger than the region obtained by dividing the image, it is determined whether only one of the first determination unit and the second determination unit belongs to a certain attribute. Compared to the above, errors in the determination can be suppressed.

It is a conceptual module block diagram about the structural example of this Embodiment. It is explanatory drawing which shows the example of the processing result by a weak-judgment module and a strong-judgment module. It is explanatory drawing which shows the example of the process result by a joint process module. It is explanatory drawing which shows the example of the process result by a labeling module. It is explanatory drawing which shows the example of the data structure of a label and a strong block flag correspondence table. It is explanatory drawing which shows the example of the block size of a strong-judgment module. It is explanatory drawing which shows the example of the block size of a weak-judgment module. It is explanatory drawing which shows the example of the relationship between a pattern block and a character. It is explanatory drawing which shows the example of the relationship between a pattern block and a character. It is explanatory drawing which shows the example of selection of the parameter used with a strong-judgment module and a weak-judgment module. It is explanatory drawing which shows the example at the time of judging with two types of parameters. It is explanatory drawing which shows the example of a judgment result. It is a block diagram which shows the hardware structural example of the computer which implement | achieves this Embodiment. It is explanatory drawing which shows the example of the image divided into layers. It is explanatory drawing which shows the example of the image divided into layers. It is explanatory drawing which shows the example of a character extraction process. It is explanatory drawing which shows the example of an error at the time of extracting as a character. It is explanatory drawing which shows the example of an error at the time of extracting as a character. It is explanatory drawing which shows the extraction process example of a pattern image. It is explanatory drawing which shows the extraction process example of a pattern image. It is explanatory drawing which shows the extraction process example of a pattern image.

  First, in order to facilitate understanding of the present embodiment, two types of technologies to which the present embodiment can be applied (technology applied after the processing according to the present embodiment is performed) will be described.

<1> The received image is divided into a picture area and a character area, and an image compression method suitable for each area is applied. This embodiment uses this embodiment in order to improve the overall image compression rate. In addition, a picture means a thing other than a character and a line drawing, and mainly has a photograph and a halftone dot image, and is a target of irreversible compression.
In general, since the amount of data of a color image is large, it is compressed and stored or transmitted. In a lossless compression method that does not give distortion to an image, a high compression rate cannot be obtained. Therefore, as a general compression image format, a JPEG (Joint Photographic Experts Group) method (more specifically, a JPEG baseline method, the following) , Simply referred to as JPEG). Since the JPEG method is a kind of lossy encoding, an image is distorted by compression. For this reason, it is difficult to obtain an extremely high compression rate.

Therefore, as a method for obtaining a higher compression rate than the conventional JPEG method, an image format having a multilayer structure as shown in Patent Document 1, Patent Document 2, and the like has been proposed.
In order to avoid the phenomenon that JPEG image distortion mainly occurs at the edge portion of an image having a large high-frequency component power, the received image is divided into a character / line drawing portion and a character / line drawing portion having a large high-frequency component power. The background portion is removed, and the background portion is compressed with JPEG at a high compression rate, and the character / line drawing portion is compressed with lossless encoding.

Among them, the technique described in Patent Document 1 employs a three-layer system.
As shown in the example of FIG. 14, the image 1410 is divided into a pattern portion 1414 and a character / line drawing portion 1412. Further, selector information 1416 is stored as to whether the pattern portion 1414 or the character / line drawing portion 1412 is selected as the final output image. An image is expressed using these three layers of data. Hereinafter, the pattern portion is a background image, and the character / line drawing portion is a foreground image.

The technique described in Patent Document 2 employs a method of compressing with a structure of a background multilevel image 1 layer and a binary image N layer. A multi-value JPEG image is used as a background image, and a binary image having color information is superimposed thereon.
The background image 1510 which is the lowest layer in the example shown in FIG. 15 is compressed in a multi-value image format such as JPEG. The other layers (binary images 1520 to 1580) are compressed using the binary image coding method. Since the color information is assigned to the image of each binary image encoding layer, the color is added and decoded at the time of decoding.

  As described above, the compression rate is increased while maintaining the image quality by dividing and encoding the pattern portion and the character portion. In the techniques described in Patent Document 1 and Patent Document 2, the region is divided by extracting a character portion from the received image. Then, the pixels other than the characters become the pattern portion.

There are various other techniques for extracting a character portion from an accepted image, such as Patent Document 3, Patent Document 4, and Patent Document 5. The character extraction methods described here also extract characters existing in the pattern.
As shown in the example of FIG. 16A, when a character actually exists in the pattern image 1610 of the image 1600, it is extracted as a character as in the image 1650 shown in the example of FIG. There is no problem.
However, as shown in the examples of FIGS. 17 and 18, even what should be the pattern in the pattern (patterns shown in the examples of FIGS. 17B and 18B) is extracted as characters. If this happens, the image quality deteriorates. Note that the pattern image 1710 of the image 1700 shown in FIG. 17A is a photographic image and is mixed with a dark line, and the design image 1810 of the image 1800 shown in FIG. It is a photographed image.
In general, the method described in Patent Document 1 sets a high foreground compression rate. In the method described in Patent Document 2, the color of an image extracted as a character is painted with one color. Therefore, when the image of the pattern part is extracted as a character, the image quality is deteriorated as compared with the case where the image is not extracted.

Therefore, preferably, in the case shown in the example of FIG. 16, the character portion should be extracted as a character, and the characters other than the character should be extracted as a design. Furthermore, in the case shown in the examples of FIGS. 17 and 18, it should not be extracted as characters.
For that purpose, it is only necessary to accurately extract the pattern portion from the received image. Desired pattern part extraction examples for the images 1600, 1700, and 1800 shown in the examples of FIGS. 16, 17, and 18 are the examples of FIGS. 19B, 20B, and 21B. The hatched portions of the pattern images 1960, 2060, and 2160 shown in FIG. The pattern image 1960 shown in the example of FIG. 19B is obtained by extracting a part other than the character part (the part “ABC”) as the pattern part.

<2> In applying image processing to the received image, an image processing method or parameter suitable for each region is used. This embodiment uses this embodiment in order to improve the image quality after image processing.
Conventionally, in image processing apparatuses such as copying machines and facsimiles, image processing suitable for each region in an image is performed to improve image quality. For example, if a character area such as a character or line drawing and a picture area such as a photo or halftone dot are mixed in the received image, the character area and the picture area are separated when the image is played back, and the character area has a resolution. From the viewpoint of image quality, it is desirable to perform processing with emphasis and to perform processing with emphasis on gradation on the pattern area.
Specifically, when an image is read by a scanner or the like, or when an image is transmitted by facsimile or the like, when printing an image in which characters, line drawings, photos, halftone dots, etc. are mixed, it is constituted by characters, line drawings, etc. Processing is performed with emphasis on resolution for the character area, and processing with emphasis on gradation is performed on the pattern area composed of photographs, halftone dots, and the like. Thereby, a high-quality reproduced image is obtained.
Therefore, even in the case of applying to such a technique, as described above, a desirable pattern portion extraction example when the images 1600, 1700, and 1800 shown in the examples of FIGS. The hatched portions of the pattern images 1960, 2060, and 2160 shown in the examples of FIGS. 19 (b), 20 (b), and 21 (b).

Next, in order to facilitate understanding of the present embodiment, a specific processing example according to the present embodiment will be shown and described.
As an example of processing according to the present embodiment, it is assumed that area determination (area division) is performed for each block.
First, area determination is performed for each block using two parameters (a strong parameter and a weak parameter). For example, it is assumed here that the pattern / character determination is performed. Note that there are various attributes such as a pattern, a photograph, a halftone dot, a character, and the like to be judged. Here, for simplicity of explanation, two attributes are a pattern and a character.
It is assumed that the strong parameter is a parameter for selecting only those that can be determined to have a pattern with high accuracy. If it is determined that the pattern is a strong parameter, the possibility of being a character is low.
The weak parameter has a lower accuracy than the strong parameter, but is a parameter for determining that it is a pattern to some extent. When it is determined that the pattern is a weak parameter, even a character may be determined as a pattern.

  A block determined by such two types of parameters is considered to have a structure as shown in the example of FIG. In the example shown in FIG. 11, one square represents one block. One block is composed of a plurality of pixels. For example, it is assumed that the rectangle is a vertical M pixel and a horizontal N pixel. Although the values of M and N are not limited, for example, they are in various ranges from several pixels (may be one pixel) to several hundred pixels. In the example shown in FIG. 11, the shaded block indicates that the block is determined to be a picture with a strong parameter (hereinafter referred to as a strong block). A vertical line block indicates that the block is determined to be a pattern with a weak parameter (hereinafter referred to as a weak block). It can be seen that the weak blocks are distributed around the strong blocks.

Usually, the pattern area has a two-dimensional extent.
When a picture area having such a two-dimensional spread is divided into blocks, there is a high possibility that a block adjacent to the picture area block is a picture area. Conversely, a block that is not adjacent to a picture area block is relatively unlikely to be a picture area.

Therefore, in the present embodiment, first, it is determined that the strong block is a picture. Next, it is determined that the weak block in contact with the strong block is a picture.
Furthermore, a weak block in contact with a block that has already been determined to be a design (hereinafter referred to as a design block) is also determined to be a design. The weak block in contact with the design block is repeatedly made into a design block. That is, after the weak block is converted into a pattern block, it is determined that the weak blocks around the weak block that has been converted into the pattern block are also patterns.
When such processing is performed on the example shown in FIG. 11, the pattern portion is determined as in the black portion shown in the example of FIG.

  By performing such processing, a weak block that is not connected to a strong block is not determined as a picture. That is, an isolated block with low accuracy that is not two-dimensionally continuous is not judged as a picture. Moreover, the weak block connected to the strong block is determined to be a picture. That is, even a block with low accuracy is determined as a pattern if it is two-dimensionally connected to a block with high accuracy.

  Processing the two-dimensionally connected blocks as described above means that a block having a larger area is used in a pseudo manner. Thereby, even if the block size is relatively small, the area determination is performed stably.

In the above description, whether or not the block is a picture is determined. However, the present invention is not limited to this. For example, it may be determined whether a block is a character or a halftone dot. It may be a photograph or not.
In the above description, “a weak block that touches a block that has already been determined to be a pattern is determined to be a pattern”. In this case, an area that is in contact with a block that has already been determined to be a pattern Therefore, high-speed determination is possible. In addition, as another processing example, blocks determined as a picture regardless of a strong block or a weak block are connected to form a block of blocks, and if there is a strong block in the block, the entire block is determined as a pattern. It is also possible to do it.
In the above description, the difference in accuracy is changed between the “strong parameter” and the “weak parameter”, but it may not be the difference in the parameter. The determination process itself, such as “strong determination process” and “weak determination process”, may be different. Also, the block size may be different, such as “wide block” and “narrow block”. In this case, “pixels determined to be a picture in a narrow block” connected to “pixels determined to be a picture in a wide block” may be determined to be a picture.

Hereinafter, an example of a preferred embodiment for realizing the present invention will be described with reference to the drawings.
FIG. 1 shows a conceptual module configuration diagram of a configuration example of the present embodiment.
The module generally refers to components such as software (computer program) and hardware that can be logically separated. Therefore, the module in the present embodiment indicates not only a module in a computer program but also a module in a hardware configuration. Therefore, the present embodiment also serves as an explanation of a computer program, a system, and a method. However, for the sake of explanation, the words “store”, “store”, and equivalents thereof are used. However, when the embodiment is a computer program, these words are stored in a storage device or stored in memory. It is the control to be stored in the device. In addition, the modules correspond almost one-to-one with the functions. However, in mounting, one module may be composed of one program, or a plurality of modules may be composed of one program. A plurality of programs may be used. The plurality of modules may be executed by one computer, or one module may be executed by a plurality of computers in a distributed or parallel environment. Note that one module may include other modules. Hereinafter, “connection” is used not only for physical connection but also for logical connection (data exchange, instruction, reference relationship between data, etc.).
In addition, the system or device is configured by connecting a plurality of computers, hardware, devices, and the like by communication means such as a network (including one-to-one correspondence communication connection), etc., and one computer, hardware, device. The case where it implement | achieves by etc. is also included. “Apparatus” and “system” are used as synonymous terms. “Predetermined” means that the process is determined before the target process, and not only before the process according to this embodiment starts but also after the process according to this embodiment starts. In addition, if it is before the target processing, it is used in accordance with the situation / state at that time or with the intention to be decided according to the situation / state up to that point.

  As shown in FIG. 1, the image processing apparatus according to the present embodiment includes a blocking module 110, a strong-determination module 120, a weak-determination module 130, a combination processing module 140, a labeling module 150, and an integrated determination processing module 160. have. The present invention relates to a technique for dividing an image for each attribute. Hereinafter, a description will be given mainly using a picture as an attribute.

  The blocking module 110 is connected to the strong-determination module 120 and the weak-determination module 130. The received image is divided into blocks. Accepting an image means, for example, reading an image with a scanner, a camera, etc., receiving an image from an external device via a communication line with a facsimile, etc., a hard disk (in addition to what is built in a computer, via a network) And the like, and the like read out the images stored in the device etc.). The image may be a binary image or a multi-value image (including a color image). One image may be received or a plurality of images may be received. Further, the contents of the image may be a document used for business, a pamphlet for advertisement, or the like. Dividing into blocks (blocking) means, for example, dividing an image received into an N × M rectangular shape here. Then, the pixel value of each block is passed to the strong-determination module 120 and the weak-determination module 130.

The strong-determination module 120 is connected to the blocking module 110, the combination processing module 140, and the integrated determination processing module 160. It is determined whether or not the block divided by the blocking module 110 belongs to the design attribute. Then, the determination result is passed to the combination processing module 140 and the integrated determination processing module 160.
The weak-judgment module 130 is connected to the blocking module 110 and the combination processing module 140. Whether or not the block divided by the blocking module 110 belongs to the picture attribute is determined by a determination method different from the determination method performed by the strong-determination module 120. Then, the determination result is passed to the combination processing module 140. The “judgment method different from the judgment method performed by the strong-judgment module 120” refers to at least one or more blocks among the blocks judged to belong to the pattern attribute by the strong-judgment module 120 (hereinafter referred to as the first block). The weak-determination module 130 may determine that the block belongs to the picture attribute (hereinafter, the second block) so that the blocks are different. Therefore, some blocks are also determined as first blocks and also as second blocks. Moreover, even if the first block number and the second block number are the same, there may be different blocks.

More specific examples of the pattern determination processing by the strong-determination module 120 and the weak-determination module 130 will be described.
First, a histogram of pixel values in a block is acquired.
If the maximum value of the histogram is equal to or less than the threshold value TH, it is determined that the block is a picture.
In the picture block, the pixel value often smoothly changes, so the histogram of the pixel value (appearance frequency for each pixel value) becomes a similar value even if the pixel value is different. Therefore, the maximum value of the histogram is a small value. Conversely, in the character block, the background pixel and the character color pixel occupy most. In this case, since the appearance frequency of the background pixels and the appearance frequency of the character color pixels are increased, the maximum value of the histogram is increased.

For example, when the received image is a grayscale image, a histogram of the pixel value itself may be acquired.
For example, when the received image is a color image, the color image is once converted into a luminance color difference color space such as YCbCr or Lab, and then the luminance signal such as the Y component or L component is converted to gray. What is necessary is just to acquire the histogram of a gray scale value by considering it as a scale value.

In the strong-judgment module 120, the threshold value TS is used as the threshold value TH in the above-described processing.
In the weak-judgment module 130, the threshold TW is used as the threshold TH in the above-described processing. The threshold value TW is a value different from the threshold value TS.
Here, since the strong-judgment module 120 is a determiner that can be confident that it is a pattern with high accuracy, the threshold value TS may be a relatively small value compared to the threshold value TW.
Further, the weak-judgment module 130 is a determiner that extracts a block that may be a picture but may be a character depending on a condition. Therefore, the threshold TW is relatively large compared to the threshold TS. It can be a value.
Further, in order to reduce the size and speed of the image processing apparatus, the strong-judgment module 120 and the weak-judgment module 130 use a common module (apparatus, program (algorithm), etc.), and the module performs processing. When performing the above, only the threshold value may be varied. That is, the strong-determination module 120 and the weak-determination module 130 may be realized by one module.

  The output of the determination result by the strong-determination module 120 and the weak-determination module 130 may be 1-bit information for each block. For example, here, it is set to 1 when it is determined to be a picture, and 0 otherwise. FIG. 2A shows an output example of the strong-determination module 120, and FIG. 2B shows an output example of the weak-determination module 130. 2A corresponds to the vertical line block (weak block) shown in the example of FIG. 11, and FIG. 2B corresponds to the shaded block (strong block) shown in the example of FIG.

  The combination processing module 140 is connected to the strong-determination module 120, the weak-determination module 130, and the labeling module 150. The determination result is received from the strong-determination module 120 and the weak-determination module 130, and the first block and the second block are combined. Then, the processing result is passed to the labeling module 150. For example, a logical sum (OR) process of the judgment results from both is performed. FIG. 3 shows an output example by the combination processing module 140.

  The labeling module 150 is connected to the combination processing module 140 and the integrated determination processing module 160. Label the output of the join processing module 140. Then, the processing result is passed to the integrated determination processing module 160. The labeling process uses a well-known algorithm that gives labels of the same value to connected pixels. Here, labeling is performed by regarding one block as one pixel. FIG. 4 shows an output example of the labeling module 150.

The integrated judgment processing module 160 is connected to the strong-judgment module 120 and the labeling module 150. Based on the determination result by the strong-determination module 120 and the determination result by the weak-determination module 130 (in this case, the processing result by the combination processing module 140 and the labeling module 150), the blocks divided by the blocking module 110 have the pattern. Determine whether it belongs to an attribute.
As a more specific example, the integrated judgment processing module 160 has a block in which the output of the strong-judgment module 120 is 1 for each label from the processing result received from the labeling module 150. Determine whether or not.

For example, in the example shown in FIG. 4, it is checked whether or not there is a pixel whose output (see FIG. 2B) is 1 in the pixel whose label is 1. When the label is 1, since such a pixel does not exist, the block corresponding to the label 1 is not determined to be a picture.
The same determination process is performed for label 2 as well. It is checked whether or not there is a pixel for which the output of the strong-decision module 120 (see FIG. 12) is 1 in the pixel of label 2. When the label is 2, since such a pixel exists, the block corresponding to the label 2 is determined to be a picture.
The blocks corresponding to the labels 3 and 4 are not determined to be pictures by the same determination process.

  As described above, in this embodiment, the design / non-design is determined for each block. Finally, the integrated judgment processing module 160 outputs a judgment result for each block as to whether it belongs to the picture attribute.

Below, the modification of this Embodiment is demonstrated.
<Block shape>
The block shape may be other than a rectangle. Further, all the blocks need not have the same shape. When a block having a shape other than a rectangle is adopted, the condition that the block A and the block B are in contact with each other is that one or more pixels in the block A and one or more pixels in the block B are in contact with each other. What is necessary is just to switch and process. In the labeling by the labeling module 150, the block is not regarded as one pixel, but may be labeled in units of one pixel at the same resolution as the received image.

<Histogram of color image>
In the strong-judgment module 120 and the weak-judgment module 130, the color image is once converted to gray scale to obtain a gray scale histogram. However, a histogram of the color value itself may be used.
If the color component is RGB, R, G, and B histograms are acquired. For other color components, such as YCbCr, histograms for Y, Cb, and Cr are acquired. If the maximum values of the respective histograms are all equal to or less than the threshold value, the pattern is determined. The threshold value may be the same value without depending on the color component, or may be a different value for each color component.

<Combination processing module 140>
In the combination processing module 140, the logical sum of the output of the strong-judgment module 120 and the output of the weak-judgment module 130 has been acquired. However, depending on the processing contents of the strong-determination module 120 and the weak-determination module 130, if the strong-determination module 120 is 1, the output of the weak-determination module 130 may logically always be 1. In such a case, the output of the weak-judgment module 130 may be used as the output of the combination processing module 140 as it is, and the combination processing module 140 is unnecessary.
This is the case where the maximum value of the histogram shown in the above embodiment is used. In this case, the connection processing module 140 is not necessary.

<Strong-Judgment Module 120 and Weak-Judgment Module 130>
In the above-described embodiment, the strong-judgment module 120 and the weak-judgment module 130 perform processing separately, but it is not always necessary to perform processing separately. In the determiner using the histogram as used in the above-described embodiment, the same block histogram may be used, and only the threshold value may be changed to determine whether the block is a strong block, a weak block, or the like.

For example, the calculation processing when the maximum value of the histogram is H and the output of the determiner in the strong-determination module 120 and the weak-determination module 130 is Y is shown.
if (H> TW) Y = 0
else if (H> TS) Y = 1
else Y = 2
This is because if H is greater than the threshold TW, Y is 0, otherwise Y is 1 if H is greater than the threshold TS, otherwise Y is 2. To do. By this arithmetic processing, the determination result of the strong-determination module 120 and the determination result of the weak-determination module 130 can be obtained at a time. In the example of the arithmetic processing described above, the LSB (least significant bit) of Y is the determination result of the weak-determination module 130, and the second bit from the bottom of Y is the determination result of the strong-determination module 120.

Further, for example, the following arithmetic processing may be performed.
if (H> TW) Y = 0
else if (H> TS) Y = 1
else Y = 3
This is because if H is greater than the threshold TW, Y is 0, otherwise Y is 1 if H is greater than the threshold TS, otherwise Y is 3. To do. Thus, the LSB of Y is the processing result by the weak-determination module 130 and the combination processing module 140, and the second bit from the bottom of Y is the determination result of the strong-determination module 120.

<Labeling Module 150 and Integrated Judgment Processing Module 160>
In the above-described embodiment, in order to simplify the description, after the labeling process is performed on the entire image, the determination process by the integrated determination process module 160 is performed. However, since raster scanning is actually performed when labeling is performed, it may be determined whether or not a strong block exists in each label while labeling.
For example, suppose that an image is labeled with a strong block being 2, a weak block being 1, and the others being 0.
Basic labeling may be performed by looking at the connected state of pixels having pixel values of 1 or more. At that time, whether or not a strong block is included in each label is entered in the label / strong block flag correspondence table 500.
The label / strong block flag correspondence table 500 is shown in the example of FIG. The label / strong block flag correspondence table 500 has a label number column 510 and a strong block flag column 520. The initial state is that all labels do not include a strong block (strong block flag = 0). Raster scanning is performed in the middle of labeling. If the pixel value in the raster scanning is 2, the strong block flag column 520 corresponding to the label is rewritten to 1. When labels are integrated, a logical sum of strong block flags corresponding to the labels to be integrated may be calculated.

<Form without labeling>
In the above-described embodiment, all the blocks connected to the strong block are extracted by the labeling module 150 performing the labeling, but it is not always necessary to do so.
The integrated determination processing module 160 uses only the determination results of the strong-determination module 120 and the weak-determination module 130 to display only the weak blocks that are in contact with the strong block and within a predetermined distance from the strong block. You may judge.

For such processing, the integrated determination processing module 160 performs the following processing ((1) to (3)).
(1) A strong block is used as a pattern block.
(2) A raster scan is performed, and a weak block in contact with the pattern block is used as a pattern.
(3) The raster scan of (2) is performed P times.
By controlling the number P of raster scans, the closeness to the strong block is controlled.

<Another form of judging device>
In the strong-judgment module 120 and the weak-judgment module 130 of the above-described embodiment, a judging device using a histogram is used, but any judging device may be used.
The strong-determination module 120 and the weak-determination module 130 determine whether or not the block belongs to the attribute of the pattern based on any one or more of the pixel value, the pixel value frequency, and the number of pixel blocks in the block. You may make it perform.
For example, the determination using the pixel value may be performed by comparing the difference value between the maximum value and the minimum value of the pixel values in the block and a predetermined threshold value. The determination using the pixel value frequency may be made by comparing the maximum value of the pixel value frequency in the block with a predetermined threshold value. Further, the determination using the number of pixel blocks may be performed by comparing the number of pixel blocks in the block with a threshold value.
You may make it perform the judgment which combined these. For example, the determination performed by comparing the difference value between the maximum and minimum pixel values in the block with a predetermined threshold value, and the comparison between the maximum pixel value frequency in the block and a predetermined threshold value. When the strong-determination module 120 determines that both determinations belong to the design attribute (logical product), the strong-determination module 120 determines that they belong to the design attribute. When it is determined that one of the determinations belongs to the attribute of the pattern (logical sum), it is determined that it belongs to the attribute of the pattern.
A more specific example is shown below.

<Another form of the judging device 1>
For example, the methods described in Patent Literature 6, Patent Literature 7, Patent Literature 8, and the like may be used as the pattern judgment devices of the strong-judgment module 120 and the weak-judgment module 130.

<Another form of the decision unit 2>
Other forms may be used as the pattern judgment unit of the strong-judgment module 120 and the weak-judgment module 130.
For example, a determination unit that does not determine small noise as shown in FIG. 17 as a character may be used.
First, a character part is extracted from the received image. As a method for extracting the character portion, a method described in Patent Literature 1, Patent Literature 2, Patent Literature 3, Patent Literature 4, Patent Literature 5, or the like may be used.
The extracted character portion is expressed as a binary image having a character portion of 1 and a non-character portion of 0. This binary image is labeled.
The labeled binary image is divided into blocks, and the number of character label types in the block is measured. The number of character label types may be regarded as the number of pixel blocks that are simply connected.
A block with a large number of character label types is a block with a lot of small noise as shown in the example of FIG. Here, the threshold value TH is set, and when the number of character label types is greater than TH, it may be determined as a picture.

<Another form of the judging device 3>
The block sizes to be processed in the strong-determination module 120 and the weak-determination module 130 may be different.
The strong-judgment module 120 increases the block size as compared to the weak-judgment module 130. If the block size is large, the accuracy of discrimination can be increased instead of the accuracy of the location of discrimination.
As the weak-judgment module 130, a smaller block size than that of the strong-judgment module 120 is adopted. Instead of increasing the position accuracy of determination, the accuracy of determination is reduced.
In order to obtain different block sizes, for example, the blocks divided by the blocking module 110 may be combined into a new size block, or the blocking module 110 may have a strong-decision module 120 and a weak-determination module 130. Therefore, it may be divided into blocks with different block sizes.
Then, the strong-determination module 120 and the weak-determination module 130 may each make a determination based on a parameter (threshold value).

When the block sizes are different, the number of output pixels of the strong-judgment module 120 and the weak-judgment module 130 will be different.
Therefore, here, processing is performed using the greatest common divisor of the block sizes of the strong-judgment module 120 and the weak-judgment module 130. The output of the strong-decision module 120 and the weak-determination module 130 may be performed with the accuracy of the block size of the greatest common divisor.
For example, the block size of the strong-judgment module 120 is 8 horizontal pixels × 8 vertical pixels (see FIG. 6A), and the block size of the weak-judgment module 130 is 6 horizontal pixels × 3 vertical pixels (see FIG. 7A). ).
Since the greatest common divisor in the horizontal direction is 2 and the greatest common divisor in the vertical direction is 1, the output block size of the strong-judgment module 120 and the weak-judgment module 130 may be 2 horizontal pixels × 1 vertical pixel. That is, the determination result is output with an accuracy of 2 horizontal pixels × 1 vertical pixel. Since the accuracy is 2 horizontal pixels × 1 vertical pixel, when the output of one block of the strong-judgment module 120 is X, all the 4 horizontal pixels and 8 vertical pixels are output as X blocks (see FIG. 6B). That's fine. When the output of one block of the weak-judgment module 130 is X, it is sufficient to output an X block (see FIG. 7B) for all three horizontal pixels and three vertical pixels.
Since the resolution of the output image created in this way is the same in the strong-judgment module 120 and the weak-judgment module 130, processing (for example, logical operation) in the combination processing module 140 or the like can be performed.
In order to be able to perform logical operations, it is not always necessary to take the block of the greatest common divisor. That is, a logical operation may be performed with a block accuracy of 1 × 1 pixel.
As described above, processing by the combination processing module 140, the labeling module 150, and the integrated determination processing module 160 may be performed as block sizes that can be logically calculated.

<Character part removal>
In the integrated determination processing module 160 of the above-described embodiment, the pattern / non-pattern is determined for each block.
In such block unit determination, it is difficult to make a pattern determination by eliminating the character portion as in the example shown in FIG.
Therefore, the following method is further used.
The integrated determination processing module 160 may determine that the block is a picture if the block belonging to the picture attribute contains the entire pixel block different from the picture attribute.
The integrated determination processing module 160 may determine that the block is not a picture when a part of a pixel block different from the picture attribute is included in the block belonging to the picture attribute.
Once it is determined that the block is a pattern, it is determined whether there is a pixel block that is a character (non-pattern). If the entire pixel block is within the block, the pixel block is Including the target block is determined to be a picture, and the pixel block is not within the block (some are in the block, but the other is not in the block. That is, the pixel block) Is present across a plurality of blocks), it is determined that the target block including a part of the pixel block is not a picture (in this case, a character).

As a more specific example, when the entire character (non-pattern) is included in the pattern block, it is determined that the character is a character in the pattern (see FIG. 8). As shown in the example of FIG. 8, even if characters protrude from one pattern block, it is determined that all characters are included in the pattern block if the protruding destination is a pattern block.
If all the characters are not included in the design block and the characters protrude from the design block, the characters are not determined to be a design (see FIG. 9).
Here, if the block size used for discrimination is sufficiently small, the character portion can be excluded and it can be determined as a picture as in the example shown in FIG.
In the above-described processing, “a pixel that is not a character” in the pattern block may be determined to be a pattern.

Furthermore, in order to make a precise determination, the integrated determination processing module 160 is based on the ratio of pixels of a pixel block that is not a picture included in a block that belongs to a picture attribute or a block that does not belong to a picture attribute. It may be determined whether or not the block is a picture.
For example, the ratio R of the pixels existing in the picture block and the pixels existing in the non-picture block among the pixels of the character pixel block is calculated as the following Expression 1.
R = number of pixels present in non-picture block / number of pixels present in picture block (Equation 1)
The ratio R is compared with a predetermined threshold, and if the ratio R is greater than the threshold, the character is determined as a non-picture. If it is smaller, the character is judged as a picture.

Alternatively, the following formula 2 may be used for the ratio R. That is, it may be a ratio between the total number of pixels of a character and the number of pixels existing in a non-picture block.
R = number of pixels present in non-picture block / total number of pixels in character part (Formula 2)
Also in this case, the ratio R is compared with a predetermined threshold value, and if it is larger than the threshold value, the character is determined as a non-picture. If it is smaller, the character is judged as a picture.

Alternatively, the following formula 3 may be used for the ratio R. That is, it may be a ratio between the total number of pixels of the character and the number of pixels existing in the picture block.
R = number of pixels present in the pattern block / total number of pixels in the character part (Formula 3)
In this case, the ratio R is compared with a predetermined threshold, and if it is larger than the threshold, the character is determined as a picture. If it is smaller, the character is judged as a non-picture.

<Determination of other attributes>
In the above-described embodiment, the determination of the picture and the non-picture is exemplified, but the attribute of the block determined by the present embodiment may be another attribute. Any attribute that is two-dimensionally distributed and easily appears as a lump can be determined stably.

<How to determine strong and weak parameters>
Here, a method for determining a strong parameter and a weak parameter will be described.
Here, as in the above-described embodiment, the maximum value of the histogram of the pixel values in the block is compared with the threshold value.
In the strong-judgment module 120, a strong parameter is compared as a threshold value. If the maximum value of the histogram is smaller than the strong parameter, it is determined as a strong block.
In the weak-judgment module 130, a weak parameter is compared as a threshold value. If the maximum value of the histogram is smaller than the weak parameter, it is determined as a weak block.
Here, a certain image is divided into blocks, and teacher data for which a picture or a non-picture is determined for each block is used. The teacher data may be given by an operation by an operator, for example.

When the maximum value of the histogram is taken on the horizontal axis and the frequency of the picture block and the non-picture block is plotted, the result is as shown in the example of FIG.
In general, the maximum value of the histogram of the pattern block is small, and the maximum value of the histogram of the non-pattern block is large.
If there is no overlap between the frequency distributions of the two attribute blocks, it is possible to make a determination with a single determiner by setting a parameter between the distributions. However, in general, the frequency distribution of the two attribute blocks overlaps.
Therefore, two parameters are set as follows.
A strong parameter is a parameter that can determine only a picture block. A block smaller than the strong parameter can be determined as a pattern block with a high probability. For example, a value that does not select any non-picture block may be a strong parameter.
The weak parameter is a parameter that can extract all the design blocks. A block larger than the weak parameter can be determined as a non-picture block with high probability. For example, a value that can select all the picture blocks may be set as a weak parameter.
As described above, it may be quite difficult to determine a value at which a non-picture block is not selected at all, or at which “all” picture blocks can be selected. Therefore, a value in which mistakes are mixed at a certain ratio may be used.
For example, a strong parameter may be a value that selects only X% or less of non-picture blocks. And the value which makes the ratio which does not select a pattern block Y% or less should just be made into a weak parameter.

  A hardware configuration example of the image processing apparatus according to the present embodiment will be described with reference to FIG. The configuration shown in FIG. 13 is configured by a personal computer (PC), for example, and shows a hardware configuration example including a data reading unit 1317 such as a scanner and a data output unit 1318 such as a printer.

  A CPU (Central Processing Unit) 1301 executes various modules described in the above-described embodiments, that is, the modules such as the blocking module 110, the strong-determination module 120, the weak-determination module 130, and the combination processing module 140. It is a control part which performs the process according to the computer program which described the sequence.

  A ROM (Read Only Memory) 1302 stores programs used by the CPU 1301, calculation parameters, and the like. A RAM (Random Access Memory) 1303 stores programs used in the execution of the CPU 1301, parameters that change as appropriate during the execution, and the like. These are connected to each other by a host bus 1304 including a CPU bus.

  The host bus 1304 is connected to an external bus 1306 such as a PCI (Peripheral Component Interconnect / Interface) bus via a bridge 1305.

  A keyboard 1308 and a pointing device 1309 such as a mouse are input devices operated by an operator. The display 1310 includes a liquid crystal display device or a CRT (Cathode Ray Tube), and displays various types of information as text or image information.

  An HDD (Hard Disk Drive) 1311 has a built-in hard disk, drives the hard disk, and records or reproduces a program executed by the CPU 1301 and information. The hard disk stores an image received by the blocking module 110, processing result data of each module, and the like. Further, various computer programs such as various other data processing programs are stored.

  The drive 1312 reads data or a program recorded on a removable recording medium 1313 such as a magnetic disk, an optical disk, a magneto-optical disk, or a semiconductor memory, and reads the data or program into an interface 1307 and an external bus 1306. , The bridge 1305, and the RAM 1303 connected via the host bus 1304. The removable recording medium 1313 can also be used as a data recording area similar to a hard disk.

  The connection port 1314 is a port for connecting the external connection device 1315 and has a connection unit such as USB and IEEE1394. The connection port 1314 is connected to the CPU 1301 and the like via the interface 1307, the external bus 1306, the bridge 1305, the host bus 1304, and the like. A communication unit 1316 is connected to a network and executes data communication processing with the outside. The data reading unit 1317 is, for example, a scanner, and executes document reading processing. The data output unit 1318 is, for example, a printer, and executes document data output processing.

  Note that the hardware configuration of the image processing apparatus illustrated in FIG. 13 shows one configuration example, and the present embodiment is not limited to the configuration illustrated in FIG. 13, and the modules described in the present embodiment are executed. Any configuration is possible. For example, some modules may be configured with dedicated hardware (for example, Application Specific Integrated Circuit (ASIC), etc.), and some modules are in an external system and connected via a communication line Alternatively, a plurality of systems shown in FIG. 13 may be connected to each other via communication lines so as to cooperate with each other. Further, it may be incorporated in a copying machine, a facsimile machine, a scanner, a printer, a multifunction machine (an image processing apparatus having any two or more functions of a scanner, a printer, a copying machine, a facsimile machine, etc.).

  In the above description of the embodiment, “greater than”, “less than”, “greater than”, and “less than (less than)” are “greater than” unless there is a contradiction in the combination. , “Smaller (less than)”, “more than”, and “less than”.

The program described above may be provided by being stored in a recording medium, or the program may be provided by communication means. In that case, for example, the above-described program may be regarded as an invention of a “computer-readable recording medium recording the program”.
The “computer-readable recording medium on which a program is recorded” refers to a computer-readable recording medium on which a program is recorded, which is used for program installation, execution, program distribution, and the like.
The recording medium is, for example, a digital versatile disc (DVD), which is a standard established by the DVD Forum, such as “DVD-R, DVD-RW, DVD-RAM,” and DVD + RW. Standard “DVD + R, DVD + RW, etc.”, compact disc (CD), read-only memory (CD-ROM), CD recordable (CD-R), CD rewritable (CD-RW), Blu-ray disc ( Blu-ray Disc (registered trademark), magneto-optical disk (MO), flexible disk (FD), magnetic tape, hard disk, read-only memory (ROM), electrically erasable and rewritable read-only memory (EEPROM), flash Includes memory, random access memory (RAM), etc. .
The program or a part of the program may be recorded on the recording medium for storage or distribution. Also, by communication, for example, a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a wired network used for the Internet, an intranet, an extranet, etc., or wireless communication It may be transmitted using a transmission medium such as a network or a combination of these, or may be carried on a carrier wave.
Furthermore, the program may be a part of another program, or may be recorded on a recording medium together with a separate program. Moreover, it may be divided and recorded on a plurality of recording media. Further, it may be recorded in any manner as long as it can be restored, such as compression or encryption.

DESCRIPTION OF SYMBOLS 110 ... Blocking module 120 ... Strong-judgment module 130 ... Weak-judgment module 140 ... Combination processing module 150 ... Labeling module 160 ... Integrated judgment processing module

Claims (7)

Dividing means for dividing the received image into a plurality of regions in an arbitrary shape ;
First determination means for determining whether the region divided by the dividing means belongs to a certain attribute;
Second determination means for determining whether or not the area divided by the dividing means belongs to the attribute by a determination method different from the determination method performed by the first determination means;
Based on the determination result by the result and the second determination means determines by said first determining means, areas divided by said dividing means comprises a third determining means for determining whether belonging to the attribute ,
The third determination means determines that the second area determined by the first determination means as not belonging to the attribute but determined as belonging to the attribute by the second determination means is the first area. An image processing apparatus characterized in that, when a part of the first area, which is an area determined to belong to the attribute by the determination means, is in contact with the attribute, the second area is determined to belong to the attribute . The first determination unit and the second determination unit are any one of a pixel value in the region, a pixel value frequency, and the number of pixel blocks that are pixel groups in which pixels having the same pixel value are connected. The image processing apparatus according to claim 1, wherein it is determined whether or not the area belongs to the attribute based on the above. The first determination means determines whether a region belongs to the attribute by comparing the maximum value of a histogram of pixel values with the first threshold value,
The second determination means uses a second threshold value different from the first threshold value used by the first determination means to determine whether a region belongs to the attribute histogram of pixel values. maximum and the image processing apparatus according to any one of claims 1 or 2, characterized in that by comparison with the second threshold. A pixel block that is a group of pixels in which pixels having the same pixel value are connected based on a ratio of pixels included in a region belonging to the attribute and pixels included in a region not belonging to the attribute There the image processing apparatus according to any one of claims 1 to 3, characterized in that it comprises a pixel block judgment means for judging whether the attribute. The pixel block determining means, if it contains the whole of the pixel block in the region belonging to the attribute, the image according to claim 4, characterized in that it is determined that該画pixel block is the attribute Processing equipment. The pixel block determining means, if it contains a portion of the pixel block in the region belonging to the attributes,該画pixel block is according to claim 4, characterized in that it is determined that not the attribute Image processing device. Computer
Dividing means for dividing the received image into a plurality of regions in an arbitrary shape ;
First determination means for determining whether the region divided by the dividing means belongs to a certain attribute;
Second determination means for determining whether or not the area divided by the dividing means belongs to the attribute by a determination method different from the determination method performed by the first determination means;
Based on the determination result by the first determination unit and the determination result by the second determination unit, the third division unit functions as a third determination unit that determines whether the area divided by the division unit belongs to the attribute. ,
The third determination means determines that the second area determined by the first determination means as not belonging to the attribute but determined as belonging to the attribute by the second determination means is the first area. An image processing program characterized by determining that the second area belongs to the attribute when part of it is in contact with the first area that is determined to belong to the attribute by the determining means .

Images