JP2008269509A - Image processing program and image processor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To extract described characters, seal impression, marks, or the like from a document image including color slurring and color burrs with high accuracy. <P>SOLUTION: An image processing method comprises a background elimination data generating part for eliminating a background part from a color image or luminance image and generating background elimination data representing parts other than a background, a contour color conversion data generating part for generating data obtained by converting the contour color of the parts other than the background into the color of pixels existing inside the contour of the parts other than the background in the parts other than the background of the color image or luminance image, and a specific object extracting part for extracting a specific object part. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  In the present invention, for example, a document such as a form is imaged with an optical character reader (OCR), a scanner, a digital camera, and the like, and a specific object such as a written character, an imprint, or a mark is extracted from the generated document image. The present invention relates to an image processing method and an image processing apparatus.

  Financial institutions and local governments use scanner devices such as OCR to improve the efficiency of document processing operations such as forms. The main functions of OCR are generation of a document image, extraction of characters in the document image, and character recognition. As the types of document images to be generated, there are binary images, luminance images, and color images.

  Since processing using a binary image has a small amount of data, processing time is reduced. However, in binary image processing, ruled lines, digit lines, guide characters, and hatching called pre-printing that are pre-printed on forms are distinguished from handwritten or printed characters that are printed later. Is difficult. For this reason, noise may be generated in the character extraction result, or a part of the extracted character may be lost, which causes a problem of erroneous character recognition.

  The processing using the luminance image is black and white grayscale image processing. In luminance image processing, the pre-printing and entry characters are distinguished from each other by using different brightness values. Therefore, when pre-printing and entry characters overlap, it is more distinctive than binary image processing. Easier. However, when the pre-printing and the brightness value of the written characters are close to each other, their discrimination accuracy is lowered.

  In the process using a color image, pre-printing and written characters can be distinguished by the difference in color, so that they can be distinguished more easily than luminance image processing. In color image processing, pre-printed colors are dropped out to extract written characters, seals, marks, and the like.

  The methods include a method of dropping out a dropout color designated before inputting a form such as “Patent Document 3”, and a case of entering a form such as “Patent Document 1” and “Patent Document 2”. There is a type of extracting a specific shape portion such as a ruled line and dropping out the same color as the color of the extracted portion.

JP2003-196592 JP2005-258683 JP 2006-134355 JP2004-336106 JP2005-18810

  In the color image processing, color misregistration occurs in images generated from OCR, scanners, and digital cameras. As a result, specific objects such as written characters and seals cannot be extracted correctly, and some preprints remain. There is a problem that a part of a specific object is missing.

  The color shift is a shift in the position of at least one of the three primary color components to be sensed, an R value that is a red component, a G value that is a green component, and a B value that is a blue component. Causes of color misregistration include chromatic aberration of the lens, sensor arrangement position, conveyance speed, and the like. In particular, in a scanner using a two-dimensional CCD such as a stand-type scanner or a digital camera, color misregistration due to chromatic aberration often occurs.

  Due to the color misregistration, a false color different from the original color of the specific target is generated in the contour portion of the specific target such as pre-printing or written characters. For example, there are cases in which red and blue false colors are generated in the outline of a black character, and light red false colors are generated in the outline of a blue ruled line. For this reason, an error occurs in color image processing for distinguishing between a written character and pre-printing depending on the color information.

  On the other hand, “Patent Document 4” attempts to eliminate chromatic aberration of the lens, and “Patent Document 5” attempts to eliminate color misregistration due to the position of the sensor.

  However, even if a method of measuring and correcting the shift amount is performed, it is difficult to completely eliminate the color shift from the image. In addition, more accurate color misregistration correction has a problem that it takes a lot of ruled line time.

  Also, in the color image processing and luminance image processing, when color blur occurs in the image, specific objects such as written characters and seal imprints cannot be correctly extracted, and a part of pre-printing may remain, There is a problem that some parts are missing.

  The color blur is a light color generated by blurring the color of a ruled line or the outline of a written character. Due to color blur, the red component, blue component, red component, brightness, saturation, hue, brightness, and other color information of preprints and written characters increase, making it difficult to distinguish between written characters and preprinted characters. Become.

  The present invention has been made in view of these problems, and an image processing method and an image processing apparatus for extracting a specific target such as written characters, seals, and marks from a document image including color misregistration and blurring with high accuracy. It is to provide.

  In order to achieve the above-mentioned object, the present invention provides a color image in an image processing method for extracting a specific object such as a letter, seal, or mark from a color image or luminance image obtained by reading a document such as a form with a scanner or a digital camera. Alternatively, the background is removed from the luminance image and background removal data generation processing for generating background removal data representing the portion other than the background, and the color information of the outline of the portion other than the background in the color image or the portion other than the background in the luminance image is obtained. It has a contour color conversion data generation process for generating data converted into color information of pixels inside a contour other than the background part, and a specific target extraction process for extracting a specific target part.

  The contour color conversion data generation processing refers to a target pixel in a color image or a luminance original image with reference to neighboring pixels that are a plurality of pixels in the vicinity thereof, and a red component, a blue component, and a red component of the target pixel. Generating low luminance color expansion luminance data in which color information such as brightness, saturation, hue, and luminance is converted into color information of a pixel having the lowest luminance value among neighboring pixels and a target pixel. .

  The specific target determination process is characterized in that ruled line extraction, specific target candidate extraction, ruled line color information and specific target color information estimation, and specific target determination are performed.

  According to the present invention, even for a color image or luminance image having a color shift or a color blur, specific objects such as pre-printing, written characters, imprints, and marks are accurately distinguished, and for example, only written characters are accurately extracted. It becomes possible to do. It is possible to accurately extract a specific target in a document image such as an imprint or a mark as well as a written character.

  Hereinafter, an image processing method and an image processing apparatus to which the present invention is applied will be described in detail with reference to the drawings.

  FIG. 2 is a diagram illustrating an embodiment of the image processing apparatus of the present invention.

  The communication device 201, the image acquisition device 202, the display device 203, the external storage device 204, the memory 205, the CPU (Central Processing Unit) 206, and the input device 207 such as a keyboard and a mouse are connected by a communication line such as a PCI bus. The image processing device 208.

  1 is stored in a storage device such as the external storage device 204 or the memory 205, and is executed by the CPU 205.

  The color image and luminance image of the document input to the CPU may be input from the image acquisition device 202 such as a scanner or OCR, the communication device 201, or may be stored in the external storage device 204.

  The result of the specific target extraction processing may be output to the display device 203, output to the outside via the communication device 201, or used for another program in the image acquisition processing device 208. Another example of the program is a program that performs character recognition.

  FIG. 27 shows an example in which a color image input from the image acquisition device 202 or the communication device 201 or stored in the external storage device 204 is displayed on the display window 2701 on the display device 203. FIG. 28 is an example in which the result of the specific target extraction process is displayed on the display window 2702 on the display 203.

  FIG. 1 is a diagram showing the configuration of a specific target extraction processing program to which the present invention is applied. The specific target extraction processing program includes an image acquisition module 101, a background removal data generation module 102, a contour color conversion data generation module 103, a specific target determination module 104, and a control module 105.

  The image acquisition module 101 performs image acquisition processing for acquiring a color image or luminance image obtained by imaging a paper document or the like with a scanner, OCR, or the like.

  The background removal data generation module 102 performs background removal data generation processing for generating background removal data from a color image or luminance image input to the CPU 206.

  For example, when a color image including ruled lines 301, digit lines 302, hatching 303 pre-printed and written characters 304 as shown in FIG. 3 is acquired, the background removal data generation module 102 enters the ruled lines, digit lines, and entries shown in FIG. Data representing the character part is generated.

  The background removal data generation process is a process for removing a background portion from an image and extracting preprinted portions and written character portions of ruled lines and digit lines. There are various methods for realizing this, and the method shown in FIG. 6 is used.

  First, in a luminance value data generation process 601, a luminance image expressed by luminance is generated from a color image expressed by RGB three primary colors (R value, G value, B value). Next, in a block division process 602, the luminance image is divided into a plurality of blocks. Finally, in the binarization process 603, for each block, binary data in which a pixel having a low luminance value is black and a pixel having a high luminance value is white is generated in the block. The binary data thus generated is background removal data in which black pixels indicate portions other than the background, as shown in FIG.

  The outline color conversion data generation module 103 receives the color image 604 and the background removal data 605 and generates outline color conversion data 1303 obtained by converting the outline color of ruled lines, digit lines, and written characters into the color of the inner part of the outline. Contour color conversion data generation processing is performed. Note that the color image 604 may be a luminance image.

  The specific target determination module 104 performs a specific target determination process for generating data representing the entered character portion shown in FIG. 5 with reference to the outline color conversion data 1303 for the background removal data 605 input to the CPU 206, An object discrimination result 706 is output.

  Here, the conventional specific target determination process will be described with reference to FIG. In the conventional specific target determination process, background removal data is input, and a specific target determination result is output with reference to a color image.

  A conventional specific target determination process is shown in FIG. First, in a ruled line extraction process 701 ', a ruled line part is extracted. In this process, the ruled line portion is extracted by extracting a portion where the black pixels in the background removal data are long and linearly connected. The result is shown in FIG.

  Next, in ruled line removal processing 702 ', ruled line removal data obtained by removing the ruled line portion from the background removal data is generated. The result is shown in FIG.

  Next, in a specific target candidate extraction process 703 ′, input character portion candidates that are candidates for the input character portion to be specified are extracted from the ruled line removal data using information on the size and position of the rectangle. The result is shown in FIG.

  Next, in the ruled line color and specific target color estimation process 704 ′, by referring to the color image 604, the ruled line part color information that is the color information of the ruled line part and the entry character candidate part color that is the color information of the entered character candidate part Estimate information.

  Next, in the identification processing 705 ′ for the specific object, it is determined whether each pixel of the black pixel portion in the background removal data is a pixel of the input character by using the ruled line partial color information and the input character candidate partial color information. To do. This process is a process for determining whether the color information of the color image belongs to the ruled line partial color information or the entry character candidate partial color information at the position of each pixel of the black pixel portion in the background removal data.

  Specifically, the following processing is performed for each pixel of the black pixel portion in the background removal data 605. In the process of a certain black pixel position (Xa, Xb) in the background removal data 605, the color information in (Xa, Xb) of the color image 604 is the ruled line output in the ruled line color and specific target color estimation process 704 ′. It is determined whether the partial color information is closer to the entry character candidate partial color information. If the color information of (Xa, Xb) is close to the ruled line partial color information, the position of (Xa, Xb) is the ruled line part, and if the color information of (Xa, Xb) is close to the input character candidate partial color information , (Xa, Xb) is determined to be an entry character portion.

  As the color information of the color image 604, R values, G values, and B values of the three primary colors of RGB may be used. Color information obtained by converting them may be luminance values, hue, saturation, and brightness of the HSV space. May be. Also, only one value may be used from among them, or a plurality of values may be used. Further, various discrimination algorithms using teacher data can be used as the discrimination method. For example, a neural network, a linear classifier, a Mahalanobis distance, or the like is used.

  Next, the specific target determination process is performed with reference to the color image 604, and the specific target determination result 706 'is output, whereby the specific target determination process ends.

  However, in the case of the conventional specific target determination process, if there is a color shift in the input color image 604, the color information in the color image 604 is also shifted, so it is closer to either the ruled line partial color information or the input character candidate partial color information. There is a possibility that the determination itself based on the color information may be deviated, and it becomes impossible to distinguish between pre-printing and written characters by the color information. For this reason, in the specific object determination result obtained by the conventional method, the determination result of the present invention, for example, the output as shown in FIG. 5 may not be obtained.

  Here, FIG. 11 is an example of an image (entry character) having a color shift. In FIG. 11, a blue false color and a red false color are generated in the outline of the originally written black character.

  FIG. 12 is also an example of an image (ruled line) having a color shift. In FIG. 12, a pale red false color is originally generated on the outline of the blue ruled line. Consider a case in which only entry characters are extracted from an image including black entry characters and blue ruled lines using color information.

  If there is no color misalignment between the entered characters and the ruled lines, only the entered characters can be extracted by the conventional specific object determination process. However, in the image having a color shift as shown in FIG. 11 and FIG. 12, since the red component exists in both the outline of the written character and the outline of the ruled line, the outline of the ruled line is generated as noise or the character The part may be missing. When color misregistration occurs in this way, there is a problem that it becomes impossible to distinguish between pre-printing and written characters by color information.

  11 and 12, the image processing apparatus to which the present invention is applied performs contour color conversion data generation processing and performs specific object determination processing with reference to the contour color conversion data.

  The outline color conversion data generation module 103 described above performs outline color conversion data generation processing. Specifically, in a portion other than the background in the color image 604, data is generated by converting the color of the contour of the portion other than the background into the color of the pixel inside the contour of the portion other than the background. That is, data is generated by converting the color information of the outlines of ruled lines, digit lines, and written characters in the color image into color information of the inner part of the outlines.

  FIG. 13 is an example of a specific processing flow of the contour color conversion data generation processing.

  In the contour color conversion data generation process, a color image 604 and background removal data 605 are input from the communication device 201, the image acquisition device 202, and the external storage device 204 via the memory 205.

  Then, in the neighborhood brightness value generation process 1301 and the low brightness color expansion process 1302, pixels in the area other than the background in the color image are selected one by one (the selected pixel is called the target pixel), and the target pixel Color information is converted. These two processes are repeated until all the pixels in the area other than the background in the color image are processed.

  In the neighborhood luminance value generation process 1302, the luminance values of the pixels in the field surrounding the vicinity of the pixel of interest (9 pixels in the field having a 3 × 3 range centered on the pixel of interest in the example of FIG. 14) are respectively generated. . Hereinafter, pixels other than the target pixel in the field are proved to be neighboring pixels. The field is not limited to 3 × 3, and may be 2 × 2 or 4 × 4, for example. The pixel of interest is not limited to the center in the field, and the field may be set to be located at any position in the field.

  Next, in the low luminance color expansion processing 1302, the color information (for example, the R value, the G value, and the B value) of the target pixel is converted into the color information of the pixel having the lowest luminance value among the target pixel and neighboring pixels. This is to convert the color information of the contour portion where the R value, G value, and B value have shifted to the color information inside the contour portion, and to artificially convert the false color to the original color information It becomes processing.

  More specifically, the luminance values of the target pixel and neighboring pixels in the field are calculated, the pixel having the lowest luminance value is extracted, and the color information of the target pixel is converted into the color information of the pixel having the lowest luminance value. To do. If the luminance value of the target pixel is the lowest, the color information of the target pixel is maintained as it is. In this way, contour color conversion data 1303, which is data in which color information is converted, is generated in the ruled line, digit line, and entered character portions in the color image 604.

  According to the contour color conversion data generation processing, for example, in the case of a written character, as shown in FIG. 14, the bright red and blue false colors in the contour portion of the written character shown in FIG. It is converted to black with a certain brightness.

  Further, according to the contour color conversion data generation process, for example, in the case of a ruled line, as shown in FIG. 15, the bright red false color in the outline of the ruled line shown in FIG. It is converted to a blue color with a low brightness.

  FIG. 16 is a diagram of a specific processing flow of the specific target determination processing in the first embodiment.

  First, background removal data is input, and a ruled line extraction process 701 is performed to extract a ruled line portion.

  Next, ruled line removal processing 702 for generating ruled line removal data obtained by removing the ruled line portion from the background removal data is performed.

  Next, a specific target candidate extraction process 703 is performed for extracting input character part candidates that are candidates for the input character part to be specified from the ruled line removal data using information on the size and position of the rectangle.

  Next, in the specific target determination process of the present invention, the RGB values of the contour color conversion data 1303 are referred to in the ruled line color / specific target color estimation process 1601 and the specific target determination process 1602.

  In the area of the outline color conversion data 1303 corresponding to the black pixel area of the background color removal data, the number of pixels having false colors generated due to color misregistration is reduced, so that the estimation of the specific target color and ruled line color is improved. As a result, the accuracy of discriminating ruled lines and specific objects is also improved.

  As described above, in the image processing apparatus 208 to which the present invention is applied, the ruled line color and the specific target color estimation 1601 and the specific target determination process 1602 can handle the entry character portion as black and the ruled line portion as blue. , It is possible to correctly determine the entered character portion.

  As described above, according to the image processing device 208, since the contour color conversion data including the RGB values after the contour color conversion processing is referred to, the entry character to be specified is extracted with high accuracy from the color image including the color shift. can do. In addition, a character recognition device that receives the input character extraction result, which is the output of the image processing device, can obtain a more accurate recognition result. In addition, although extraction of written characters is used as an example, extraction with imprints and marks is similarly possible with high accuracy.

  Next, another embodiment of the present invention will be described.

  As illustrated in FIG. 17, the specific target determination unit 104 may perform specific target extraction processing in which a specific target is determined using only ruled line color estimation.

  In the process shown in FIG. 17, in the ruled line color estimation process 1701, only the color information of the ruled line is estimated with reference to the outline color conversion data. Next, in the ruled line color part removal processing 1702, the ruled line color part is removed from the background removal data 605 using the ruled line color information to determine the entry character part to be specified.

  As shown in FIG. 18, (specific target extraction processing in which the specific target determination unit 104 performs the specific target determination process using only the estimation of the specific target color may be used.

  In the process shown in FIG. 18, in the specific target color estimation process 1801, only the color information of the specific target candidate is estimated with reference to the contour color conversion data 1303. Next, in a specific target color portion extraction process 1702, a character portion to be specified is extracted from the background removal data 605 using the color information of the specific target.

  As illustrated in FIG. 19, the specific target determination unit 104 may perform specific target determination processing using clustering to perform specific target extraction processing.

  The processing shown in FIG. 19 performs determination using only the color information of the portion other than the background without using the result of ruled line extraction. First, in the clustering process 1901, the contour color conversion data 1303 of the portion other than the background is clustered. For clustering, R values, G values, and B values of the three primary colors of RGB may be used, or color information obtained by converting them may be used, for example, luminance values, hue, saturation, and brightness of the HSV space. Also, only one value may be used from among them, or a plurality of values may be used. Clustering methods include the k-means method, the region expansion method, and discriminant analysis.

  Next, in a target cluster selection process 1902, a target cluster is selected from a plurality of clusters obtained by clustering. There are various selection methods. For example, there is a method of selecting a cluster having a high luminance value.

  Then, in the extraction 1903 of the specific target cluster color part, the characters having the color information of the selected cluster are extracted from the black pixel part of the background removal data, thereby extracting the character to be specified.

  The specific target extraction processing program shown in FIG. 1 may further include a specific target extraction process in which a color misregistration correction module 2001 is added.

  This specific target extraction processing program has a configuration as shown in FIG. 20, and performs the same processing as the embodiment shown in FIG. 1 except for the processing described below.

  The color misregistration correction module 2001 executes color misregistration correction processing. In the color misregistration correction process, the color misregistration correction data, which is data in which the color misregistration is reduced by shifting or enlarging or reducing the R value, G value, and B value of the color image 604 acquired by the document image acquisition process. Is generated.

  In the configuration shown in FIG. 1, the data input to the background removal data generation process and the outline color conversion data generation process uses the color image 604, whereas in the embodiment of FIG. Data input to the processing and the contour color conversion data generation processing is color misregistration correction data. Thereby, even in an image having a large amount of color misregistration, it is possible to accurately extract a specific object such as a written character.

  The specific target extraction process may be made by adding a specified color acquisition module 2101 to the configuration of the specific target extraction processing program shown in FIG.

  This specific target extraction process is configured as shown in FIG. 21, and performs the same process as the embodiment shown in FIG. 1 except for the process described below.

  The designated color acquisition module performs designated color acquisition processing. In the specified color acquisition process, specified extraction target color information 2203 that is a color specified as a specific target to be extracted is acquired. Examples of the designation extraction target color information include those designated by the user in advance with a program and those inputted from an input device such as a keyboard and a mouse. As the color information, RGB R value, G value, and B value may be used, or color information obtained by converting them may be, for example, luminance value, hue, saturation, and brightness of HSV space. Also, only one value may be used from among them, or a plurality of values may be used. Further, it may be a value indicating one color or a range of color values.

  Then, the specific target determining process is a process including the specified extraction target color information 2203 in the input as shown in FIGS.

  FIG. 22 shows the specific target determination 2201 using the result of the specified extraction target color information 2203 and the ruled line color and specific target color estimation 1601 in the specific target determination processing of FIG.

  FIG. 23 identifies a specific target cluster 2301 using the specified extraction target color information 2203 in the specific target determination processing using the clustering 1901 in FIG. 19.

  The embodiments described so far are effective not only for the problem of color misregistration but also for the problem of color blur. FIG. 24 is an example of color blur in which the outline portion of a written character is a light color. When contour color conversion data generation processing is performed on the image of FIG. 24, contour color conversion data 1303 shown in FIG. 25 is generated. In the contour color conversion data 1303, a blurred light color portion in the input color image 604 is converted into a dark color. As a result, it is possible to extract a specific target with high accuracy even for an image having a color blur.

  The embodiments described so far are effective not only when a color document is input but also when a luminance image in which color blur occurs is input. When a luminance image is input, this can be dealt with by performing the contour color conversion data generation process as shown in FIG. 26 in the embodiment of FIG.

  In FIG. 26, the luminance image 2604 and the background removal data 605 are input, and the luminance image low luminance color expansion processing 2601 is performed pixel by pixel in a portion other than the background in the luminance image 2604. Then, contour color conversion data 1303 that is data obtained by converting the luminance value in the luminance image 2604 is generated in portions other than the background, that is, ruled lines, digit lines, and written character portions.

  In the low luminance color expansion processing 2601 of the luminance image, the luminance value of the pixel having the lowest luminance value among the target pixel and the neighboring pixels is converted into the luminance value of the target pixel.

The figure showing the structure of specific object extraction processing. The figure showing an image processing apparatus. An example of a color image. Background removal data. Determination result of specific target. The figure showing the example of a background removal data generation process. The figure showing the conventional specific object discrimination | determination process. Ruled line extraction result. Ruled line removal result. Specific target candidate extraction results. An example of color misalignment of written characters. Example of ruled line color shift. The figure showing the example of an outline color conversion data generation process. The figure showing the outline color conversion data generation process of the image of FIG. The figure showing the outline color conversion data generation process of the image of FIG. The figure showing specific object discrimination processing. The figure showing the specific object discrimination | determination process using only the ruled line color estimation. The figure showing the specific object discrimination | determination process using only estimation of a specific target color. The figure showing the specific object discrimination | determination process using clustering. The figure showing the structure of the specific object extraction process program which added color misregistration correction. The figure showing the structure of the specific object extraction process program provided with the extraction object color designation | designated function. The figure showing the subordinate specific object discrimination | determination processing program containing a designation | designated extraction object color. The figure showing the specific object discrimination | determination process using the clustering containing the designation | designated extraction object color. The figure showing the example of color blur. The figure showing the mode of the outline color conversion data generation process with respect to the image with a blur. The figure showing the example of the outline color conversion data generation process of luminance image input. Example of color image display. A display example of the discrimination result of a specific target.

Explanation of symbols

DESCRIPTION OF SYMBOLS 101 ... Image acquisition module 102 ... Background removal data generation module 103 ... Contour color conversion data generation module 104 ... Specific object discrimination module 105 ... Control module

Claims (15)

Means for removing the background from the input image information and generating background removal data representing an area other than the background;
Means for generating color conversion data obtained by converting color information of a pixel of a contour relating to a specific object into color information of a pixel inside the contour in a region other than the background of input image information;
Means for storing the background removal data and the color conversion data;
An image processing apparatus comprising: a specific target determining unit that selects a specific target candidate from the background removal data and outputs the specific target portion with reference to the color conversion data.   The image processing apparatus according to claim 1, wherein the image information is color image information or luminance information of an image. In the means for generating the color change data,
Selecting a field including a plurality of pixels from background removal data representing an area other than the background;
Generating a luminance value of the pixels in the field;
2. The image processing apparatus according to claim 1, wherein color conversion data is generated by converting color information of a target pixel in the field into color information of a pixel having the lowest luminance value among the pixels in the field.   The image processing apparatus according to claim 1, wherein the image information is any one of information stored in an external storage device, information acquired by an image acquisition device, or information input from a communication device. The means for generating the color conversion data includes:
Set the pixel of interest on the image information, refer to neighboring pixels that are pixels in a field surrounding the neighborhood of the pixel of interest,
2. The image processing according to claim 1, wherein the color information of the pixel of interest generates low-luminance color expansion luminance data in which color information of a pixel having the lowest luminance value among neighboring pixels and the pixel of interest is converted. apparatus.   The image processing apparatus according to claim 1, wherein the color information is at least one of a red component, a blue component, a red component, lightness, saturation, hue, and luminance. Furthermore,
Means for extracting ruled line information from the background removal data;
Means for removing ruled line information extracted from the background removal data;
2. A specific target determining unit that selects a specific target candidate from the background removal data from which the ruled line information has been removed, and outputs a specific target portion with reference to the color conversion data. Image processing apparatus. Using a processing device comprising storage means, image acquisition means, calculation means, and display means,
Removing the background from the image information input from the image acquisition means, and generating background removal data representing a region other than the background;
A step of generating color conversion data obtained by converting color information of a pixel of a contour related to a specific object into color information of a pixel inside the contour in a region other than the background of input image information;
Storing the background removal data and color conversion data in a storage means;
A specific target determining step of selecting a specific target candidate from the background removal data stored in the storage unit, referring to the color conversion data, and outputting the specific target portion to the display unit. Image processing method.   The image processing method according to claim 8, wherein the image information is color image information or luminance information of the image. In the step of generating the color conversion data,
Select a field including a plurality of pixels from background removal data representing image information other than the background,
Generate brightness values for the pixels in the field,
9. The image processing method according to claim 8, wherein the color conversion data is generated by converting the luminance value of the pixel of interest in the field into the lowest luminance value of the pixels in the field.   9. The image processing method according to claim 8, wherein the image information is any one of information stored in an external storage device, information acquired by an image acquisition device, or information input from a communication device. In the step of generating the color conversion data,
For a target pixel in the field, refer to a neighboring pixel that is a pixel other than the target pixel in the field,
9. The image processing according to claim 8, wherein the color information of the pixel of interest generates low-luminance color expansion luminance data in which color information of a pixel having the lowest luminance value among neighboring pixels and the pixel of interest is converted. Method.   The image processing method according to claim 8, wherein the color information is at least one of a red component, a blue component, a red component, lightness, saturation, hue, and luminance. Furthermore,
Extracting ruled line information from the background removal data;
Removing ruled line information extracted from the background removal data;
9. The specific target determination step of selecting a specific target candidate from the background removal data from which the ruled line information is removed, and outputting the specific target portion with reference to the color conversion data. Image processing method. Computer for image processing,
Means for removing the background from the image information input from the image acquisition means and generating background removal data representing an area other than the background;
Means for generating color conversion data obtained by converting color information of a pixel of a contour relating to a specific object into color information of a pixel inside the contour in a region other than the background of input image information;
Means for storing the background removal data and color conversion data in a storage means;
Means for selecting a specific target from the background removal data stored in the storage means, and by referring to the color conversion data and outputting a specific target portion to the display means;
Image processing program to function as

Images