JP2009182655A - Image processing apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image processing apparatus generating a PDF file which performs a character recognition processing in a high image quality and a high compression rate. <P>SOLUTION: Out of characters extracted by an extraction means for extracting characters from an image, partial area characters included in partial areas of the image expressed only in a single color and lighter in color than the corresponding partial area are retrieved. For characters not retrieved, first binary image information consisting of only characters having the same color as that of the characters not retrieved is generated. For the partial areas including the retrieved partial area characters, when partial area images from which the partial area characters have been eliminated can be expressed only in a single color, second binary image information consisting of the partial area images and characters having the same color as that of the partial area images is generated. Using the generated first binary image information and second binary image information, multivalued image information is generated by erasing the areas extracted as the binary images using the peripheral colors of those areas, and character information indicating only the characters extracted by the extraction means is generated. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an image processing apparatus.

  Conventionally, PDF (Portable Document Format) has been used as an electronic file format. This PDF is also used when, for example, an image read by a scanner is digitized.

As a technique related to such a PDF, Patent Document 1 discloses a technique for extracting a character region from a multi-value image, generating a character-free multi-value image and a character image, and obtaining a highly compressed PDF. Specifically, when the reverse character area (outline character) is also determined to be a character, a method of extracting the white character portion and binarizing it as a white character, and one threshold for dividing the lower character and the character portion A method is disclosed in which a value is detected and reversed if the pattern is reversed.
JP 2002-077633 A

  In the PDF, when character recognition is performed, a binary image or a multi-value image in which character portions are filled with a background color is performed. Higher image quality and higher compression can be achieved by filling the area surrounding the character with the background color than filling it with color. However, when a non-character part is binarized and the character part is left in a multi-valued image, the character recognition process (OCR process) using character (binary image) pixels cannot be correctly processed in the secondary use. There was a problem.

  In view of the above problems, an object of the present invention is to provide an image processing apparatus that generates a PDF file that can execute character recognition processing with high image quality and high compression.

  In order to achieve the above object, the invention of claim 1 reads an image drawn on a recording medium, generates image information indicating the image, and an image generated by the image information generation unit. Extraction means for extracting characters in the image from the information, and the color of the partial area character that is a character included in the partial area of the image represented by only a single color among the characters extracted by the extraction means A search means for searching for a character that is lighter than the color density of the partial area, and a character that is not searched for by the search means is a first character composed of only characters of the same color as the character. A first binary image generating means for generating binary image information; and a partial area image indicating an area excluding the partial area character in the partial area including the partial area character searched by the search means. A second binary image means for generating second binary image information composed only of characters of the same color as the partial area image and the partial area image when the color can be expressed only by the color; An area extracted as the binary image using the first binary image information generated by the binary image generation means and the second binary image information generated by the second generation means. Multi-value image information generating means for generating multi-value image information erased by using the peripheral color of the area, character information generating means for generating character information indicating only the characters extracted by the extracting means, An image processing apparatus comprising: an association unit that associates character information, the first binary image information, the second binary image information, and the multi-value image information as information indicating the image.

  In order to achieve the above object, the invention according to claim 2 reads an image drawn on a recording medium, generates image information indicating the image, and an image generated by the image information generation unit. Extraction means for extracting characters in the image from information, and first binary image generation means for generating first binary image information composed only of characters having the same color as the characters extracted by the extraction means And the density of the color of the partial area character that is included in the partial area of the image expressed by only a single color among the characters extracted by the extracting means is compared with the density of the color of the partial area. A partial area image showing an area excluding the partial area character in the partial area including the partial area character searched by the searching means and the partial area character searched by the search means can be expressed only by the single color In A second binary image generating unit configured to generate second binary image information including only the partial region image and characters of the same color as the partial region image; and the first binary image generating unit. The area extracted as a binary image using the first binary image information and the second binary image information generated by the second generation means uses a peripheral color of the area. Multi-value image information generating means for generating multi-value image information erased in step 3, and third binary value for generating third binary image information indicating an image including only the partial area characters searched by the search means Image information generating means, character information generating means for generating character information indicating only the characters extracted by the extracting means, the character information, the first binary image information, the second binary image information, The third binary image information, and the Image processing apparatus having a value image information, and means associated with associating the information indicating the image.

  The invention according to claim 3 is the invention according to claim 1 or 2, wherein the first compression means for compressing the multi-value image information, the character information, the first binary image information, and the And second compression means for compressing at least one piece of the second binary image information.

  According to the first aspect of the present invention, it is possible to provide an image processing apparatus that generates a PDF file that can execute character recognition processing with high image quality and high compression.

  According to the second aspect of the present invention, it is possible to provide an image processing apparatus that generates a PDF file that can execute character recognition processing with high image quality and high compression.

  According to the invention of claim 3, since at least one of the character information, the first binary image information, and the second binary image information can be compressed, the entire information amount can be compressed. .

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the following description, an embodiment in which the image processing apparatus is applied to a printer will be described. A recording medium such as paper is expressed as paper.

  First, the hardware configuration of the printer 10 will be described with reference to FIG. The printer 10 includes a central processing unit (CPU) 40, a flash memory 42, a random access memory (RAM) 44, a non-volatile memory (NVM) 46, a user interface (UI) 48, and a hard disk drive (HDD). ) 50, a printer engine 52, a scanner engine 54, and a bus 56.

  The CPU 40 governs the overall operation of the printer 10, and the processing shown in the flowchart described later is executed by the CPU 40. The flash memory 42 stores a program developed in the RAM 44, a boot program that operates at startup, and the like.

  The RAM 44 is a storage device in which an OS (Operating System), programs, image information, and the like are expanded. The NVM 46 stores setting values and the like related to the printer 10.

  The printer engine 52 is an engine that prints an image on paper. The scanner engine 54 reads an image drawn on a sheet and generates image information indicating the image.

  The HDD 50 is a storage device that stores image information and the like. The UI 48 is used when the user inputs operations of the printer 10 and information. The bus 56 is used when information is exchanged.

  It should be noted that a NIC (Network Interface Card) for connecting to a network, a driver thereof, a USB device, an interface for connecting to a telephone line, or the like may be added to the configuration described above. .

  Hereinafter, image processing according to the present embodiment will be described using a flowchart. First, image processing (part 1) will be described with reference to FIG. This image processing shows processing for reading the paper shown in FIG. 3 (hereinafter referred to as a document) and generating the images shown in FIGS. The flowchart will be described with these drawings.

  First, in step 101, an image drawn on a sheet (original in FIG. 3) is read by the scanner engine 54, and image information indicating the image is generated. In the document shown in FIG. 3, the characters A and C are the same color with a single color. The partial areas A and B are a single color and the same color. Furthermore, the character B is a white character.

  In the next step 102, characters in the image are extracted. In the manuscript of FIG. 3, three “ABC” characters A to C are extracted. In the next step 103, it is determined whether or not there is a single color partial region including the extracted character. In the document of FIG. 3, it is determined that there are two partial areas as shown in FIG.

  If a negative determination is made in step 103, character information is generated in step 111, compression processing is performed in step 112, and image information and character information are associated in step 113. Here, the character information is binary image information that is optimal as an OCR (Optical Character Recognition) processing target, and is information that indicates, for example, a character-only image as shown in FIG. The compression process indicates a process of compressing, for example, in the JPEG format when compressing a multi-valued image, and compressing in a MMR format, for example, when compressing a binary image.

  Further, associating image information with character information means creating a PDF file in which character information is associated as information indicating characters drawn in the image information. This association is used in the following description in the sense of creating an associated PDF file.

  Returning to the processing of step 103, in the case of FIG. 3, as shown in FIG. 3, since there are single-color partial areas A and B including the extracted characters, an affirmative determination is made in this step.

  If the determination in step 103 is affirmative, in step 104, a character having a character color density lower than the partial region color density is searched. In the case of the original shown in FIG. 3, since the color density of the character B is lower than the color density of the partial area in the partial area including the character B, the character B included in the partial area is searched. The density in the present embodiment indicates, for example, the sum of weighted values of R (Red), G (Green), and B (Blue), and the larger the sum, the lower the density. good.

  In the next step 105, in the case of a character that has not been searched, first binary image information composed only of characters having the same color as the character is generated. In the case of FIG. 3, the characters that have not been searched are the characters A and C. Therefore, the first binary image is the image shown in FIG.

  In the next step 106, if the partial area image including the searched partial area character and representing the area excluding the partial area character can be expressed only in a single color, the partial area image and the same color as the partial area image are displayed. Second binary image information composed only of characters is generated. Since the searched character is the character B, and the partial region image including this character is the partial region B, the second binary image information is an image shown in FIG.

  In the next step 107, multi-value image information indicating an image not included in any of the image indicated by the first binary image information and the image indicated by the second binary image information is generated. Since characters A and C are shown in FIG. 4 (B) and character B is shown in FIG. 4 (C), the image shown in FIG. 3 is included in both FIG. 4 (B) and FIG. 4 (C). The unsuccessful image is shown in FIG.

  In the next step 108, character information is generated. This character information is information indicating only the characters extracted in step 102, and in the case of the original shown in FIG. 3, the image shown in FIG. 5 is obtained. This is an image to be subjected to OCR processing.

  In the next step 109, the above-described compression processing is performed. In this compression processing, the multi-valued image information shown in FIG. 4A is compressed (for example, JPEG format), and the character image information, the first binary image information, and the second binary image information are At least one piece of information is compressed (for example, MMR format).

  In the next step 110, the first binary image information, the second binary image information, the multi-value image information, and the character information are associated. This process is a process for indicating that the first binary image information, the second binary image information, and the multi-value image information are layers of one image. Character information can be associated as information indicating the character drawn on the image.

  In this way, what can be expressed as a binary image is generated as a binary image, and what is not can be generated as a multi-valued image, so that a high-quality image can be obtained. Further, high compression can be realized by reducing the information amount of the multi-valued image. Furthermore, since character information is generated, there is no need to perform OCR processing again.

  Next, image processing (part 2) will be described with reference to FIGS. This image processing shows processing for reading the paper shown in FIG. 8 (hereinafter referred to as a document) and generating character information indicating the images shown in FIGS. 9 and 10 and the image shown in FIG. It is. The flowchart will be described with these drawings.

  First, in step 201, an image drawn on the original in FIG. 8 is read to generate image information indicating the image. In the document shown in FIG. 8, the characters A and C are the same color with a single color. Characters A, B, and C have a single color, but are all different colors. Further, unlike the original of FIG. 3, the character B is not white and has a single color different from the partial area. Furthermore, the partial areas A and B are a single color and the same color.

  In the next step 202, characters in the image are extracted. In the manuscript of FIG. 8, three “ABC” characters A to C are extracted. In the next step 203, it is determined whether or not there is a single color partial area including the extracted character.

  If a negative determination is made in step 203, the character information shown in FIG. 5 is generated in step 206, the compression processing described above is performed in step 207, and the image information and character information are associated in step 208.

  Returning to the processing of step 203, in the case of FIG. 8, since there is a single color partial region including the extracted character as shown in FIG. 8, an affirmative determination is made in this step.

  If the determination in step 203 is affirmative, in step 204, a character having a character color density lower than the partial region color density is searched. In the case of the manuscript of FIG. 8, since the density of the color of the character B is lower than the density of the color of the partial area in the partial area including the character B, the character B included in the partial area is searched.

  In the next step 205, first binary image information composed only of characters of the same color as the extracted characters is generated. In the case of this embodiment, the first binary image is the image shown in FIGS. 9B and 10B.

  The continuation of this flowchart will be described with reference to FIG. In step 209, if the partial area image including the searched partial area character and the partial area image indicating the area excluding the partial area character can be expressed by only a single color, the partial area image and the character of the same color as the partial area image 2nd binary image information comprised only by this is produced | generated.

  In the case of FIG. 8, the partial area image including the searched partial area character is an image showing the partial area B, and the character having the same color as the partial area image is the character C. Accordingly, the second binary image information is the image shown in FIGS. 9C and 10C.

  In the next step 210, it is determined whether or not to generate the third binary image information indicating an image including only the searched partial area characters. If a negative determination is made in step 210, multi-value image information indicating an image that is not included in either the image indicated by the first binary image information or the image indicated by the second binary image information is obtained in step 216. Generate. The image indicated by the multi-value image information is the image shown in FIG.

  Thereafter, in step 217, the character information shown in FIG. 5 is generated, the compression processing described above is performed in step 218, and in step 219, the first binary image information, the second binary image information, and the multi-value image information. Then, the process ends after associating the character information.

  On the other hand, when an affirmative determination is made in step 210, in step 211, third binary image information indicating an image including only the searched partial region characters is generated. Since the original in FIG. 8 shows an image including only the letter B, the image indicated by the third binary image information is the image shown in FIG.

  In the next step 212, multiple images indicating images that are not included in any of the image indicated by the first binary image information, the second binary image information, and the image indicated by the third binary image information. Value image information is generated. The image indicated by the multi-value image information is the image shown in FIG.

  After that, in step 213, the character information shown in FIG. 5 is generated, and in step 214, the above-described compression processing is performed. In step 219, the first binary image information, the second binary image information, the third 2 The process ends after associating value image information, multi-value image information, and character information.

  The compression process compresses the multi-value image information, and at least one of the character image information, the first binary image information, the second binary image information, and the third binary image information. This is a process of compressing one piece of information (for example, MMR format).

  As described above, in the image processing (part 2), since a binary image that can be expressed as a binary image is generated and a non-image is generated as a multi-valued image, a high-quality image can be obtained. Further, high compression can be realized by reducing the information amount of the multi-valued image. In addition, since character information is generated, it is not necessary to perform OCR processing again.

  The processing flow of each flowchart described above is an example, and the processing order may be changed, new steps may be added, or unnecessary steps may be deleted without departing from the scope of the present invention. Needless to say, you can.

FIG. 2 is a diagram illustrating a hardware configuration of a printer. It is a flowchart which shows image processing (the 1). FIG. 3 is a diagram illustrating a document (part 1). It is a figure which shows the image which multi-value image information and 1st, 2nd binary image information show. It is a figure which shows the image which shows character information. It is a flowchart which shows an image process (1 of the 2). It is a flowchart which shows an image process (the 2 of 2). It is a figure which shows a manuscript (the 2). It is a figure which shows the image which multi-value image information and 1st, 2nd binary image information show. It is a figure which shows the image which multi-value image information and 1st, 2nd, 3rd binary image information show.

Explanation of symbols

10 Printer 40 CPU
44 RAM
52 Printer Engine 54 Scanner Engine

Claims (3)

Image information generating means for reading an image drawn on a recording medium and generating image information indicating the image;
Extraction means for extracting characters in the image from the image information generated by the image information generation means;
Of the characters extracted by the extracting means, the density of the color of the partial area character that is a character included in the partial area of the image expressed only in a single color is compared with the color density of the partial area. A search means for searching for thin characters;
First binary image generation means for generating first binary image information composed only of characters having the same color as the character in the case of a character not searched by the search means;
In the partial area including the partial area character searched by the search means, the partial area image and the partial area can be expressed only when the partial area image indicating the area excluding the partial area character can be expressed by the single color. Second binary image means for generating second binary image information composed only of characters of the same color as the image;
As the binary image, the first binary image information generated by the first binary image generating unit and the second binary image information generated by the second generating unit are used. Multi-value image information generating means for generating multi-value image information in which the extracted region is erased by using the peripheral color of the region;
Character information generating means for generating character information indicating only the characters extracted by the extracting means;
Association means for associating the character information, the first binary image information, the second binary image information, and the multi-value image information as information indicating the image;
An image processing apparatus. Image information generating means for reading an image drawn on a recording medium and generating image information indicating the image;
Extraction means for extracting characters in the image from the image information generated by the image information generation means;
First binary image generation means for generating first binary image information composed only of characters of the same color as the characters extracted by the extraction means;
Of the characters extracted by the extracting means, the density of the color of the partial area character that is a character included in the partial area of the image expressed only in a single color is compared with the color density of the partial area. A search means for searching for thin characters;
In the partial area including the partial area character searched by the search means, the partial area image and the partial area image when the partial area image indicating the area excluding the partial area character can be expressed only by the single color. Second binary image generation means for generating second binary image information composed only of characters of the same color as
Extracted as a binary image using the first binary image information generated by the first binary image generation means and the second binary image information generated by the second generation means Multi-value image information generating means for generating multi-value image information obtained by erasing the generated region by using the peripheral color of the region;
Third binary image information generating means for generating third binary image information indicating an image including only the partial area character searched by the search means;
Character information generating means for generating character information indicating only the characters extracted by the extracting means;
Association means for associating the character information, the first binary image information, the second binary image information, the third binary image information, and the multi-value image information as information indicating the image;
An image processing apparatus. First compression means for compressing the multi-value image information;
Second compression means for compressing at least one of the character information, the first binary image information, and the second binary image information;
The image processing apparatus according to claim 1, further comprising: