JP2005301662A - Image dictionary forming device, encoding device, image dictionary forming method, and program thereof - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an encoding device which performs encoding by forming an image dictionary realizing high encoding efficiency and applying the image dictionary. <P>SOLUTION: An image processor 2 acquires a character image included in an input image and the character code of the character image, reads one or more font images corresponding to the acquired character code out of a font database, and compares the read font images with the character image in the input image to decide degrees of coincidence. Further, the image processor selects a font image to be registered as an image pattern out of those font images based upon the decided degrees of coincidence, gives an index to the selected font image, and registers it as an image dictionary. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an encoding apparatus that creates an image dictionary that associates an image pattern constituting an input image with identification information of the image pattern, and applies the created image dictionary to an encoding process.

For example, Patent Document 1 discloses a character area encoding method that includes a common font database on the encoding side and the decoding side and encodes character codes, font types, and the like.
JP-A-10-178638

  The present invention has been made from the above-described background, and an object of the present invention is to provide an encoding apparatus that creates an image dictionary that realizes high encoding efficiency and encodes the image dictionary using the image dictionary.

[Image dictionary creation device]
In order to achieve the above object, an image dictionary creating apparatus according to the present invention is provided by a template providing unit that provides a typical character image as a template image, a character image included in an input image, and the template providing unit. A matching determination unit for determining a matching degree with a template image to be obtained, and a template image provided by the template providing unit based on the matching degree determined by the matching determination unit. And a type determining unit for adding identification information for identifying each image pattern to the image pattern determined by the type determining unit.

  Preferably, the template providing means provides a font image as the template image, and the match determination means compares the provided font image with a character image included in the input image, and determines the degree of match. judge.

  Preferably, the template provision unit further includes a font storage unit that stores at least one font image in association with each character code, and an information acquisition unit that acquires a character code of the character image included in the input image. The means reads out a font image corresponding to the character code acquired by the information acquisition means from the font storage means, provides the read font image as the template image, and the match determination means includes the provided font The degree of coincidence is determined by comparing the image and the character image included in the input image.

  Preferably, when the template providing unit provides a plurality of font images corresponding to a character code, the match determination unit includes a plurality of provided font images and a character image included in the input image. In comparison, the degree of coincidence is determined for each font image, and the type determining means selects, as the image pattern, the font image that most closely matches the character image based on the degree of coincidence determined for each font image. .

  Preferably, an information acquisition unit that acquires character region information indicating a region of each character image in the input image, and a character image that extracts a character image from the input image based on the character region information acquired by the information acquisition unit And a matching unit that compares the character image extracted by the character image extracting unit with the template image provided by the template providing unit to determine the degree of matching.

  Preferably, the coincidence determination unit compares the character image extracted by the character image extraction unit and the template image provided by the template providing unit at a plurality of relative positions, and determines the degree of coincidence, The image forming apparatus further includes area correction means for correcting the character area information of each character image based on the degree of matching determined at each relative position by the matching determination means.

  Preferably, the pattern storage means for storing the image patterns sequentially determined by the type determination means, the image pattern stored by the pattern storage means and the newly input character image are compared, and either A registration control means for prohibiting the determination of the image pattern for the character image when the image pattern matches the character image, and the type determination means is responsive to an instruction from the registration control means. When the character image input to the character image matches the image pattern stored in the pattern storage means, the template image provided for the character image is not registered in the pattern storage means, and the input character image is If no image pattern matches, the template image provided for this character image is used as the image pattern. It is registered in the over down the storage means.

  The image dictionary creation device according to the present invention includes an information acquisition unit that acquires character identification information for identifying a character represented by a character image included in an input image, and character identification information acquired by the information acquisition unit. Based on the character classification means for classifying the character images included in the input image into a plurality of character image groups, template providing means for providing a typical character image as a template image, and classifying the character images into character image groups by the character classification means Each of the generated character images is compared with a template image provided for each character image group by the template providing means, and a matching determination means for determining the degree of matching respectively, and provided by the template providing means Based on template images and the degree of matching determined for these character images and template images Has a type determining means for determining an image pattern, the image pattern determined by the type determining unit, and identification information assigning means for assigning identification information for identifying each image pattern.

  Preferably, the type determining means determines the appearance frequency in each character image group of the template image that most closely matches each character image based on the degree of matching determined by the match determining means, and the determined appearance frequency Based on the above, one or more image patterns are determined for a plurality of character images classified into the same character image group.

  Preferably, the match determination means compares a plurality of character images classified into each character image group by the character classification means with a template image, determines a matching degree for each area included in the character image, The type determining unit determines a plurality of image patterns constituting the character image based on the matching degree determined for each region by the matching determining unit and the template image.

[Encoding device]
In addition, the encoding apparatus according to the present invention includes a template providing unit that provides a typical character image as a template image, a character image included in the input image, and a matching degree between the template image provided by the template providing unit. And a type determining unit for determining a template image provided by the template providing unit as a typical image pattern in the input image based on the degree of matching determined by the match determining unit. , Identification information providing means for assigning identification information for identifying each image pattern to the image pattern determined by the type determining means, and character image data included in the input image. Character area information indicating the area in the input image and the image corresponding to this character image And an encoding means for encoding by replacing the pattern identification information.

  Preferably, the matching determination unit compares the character image included in the input image with the template image provided by the template providing unit at a plurality of relative positions, determines the degree of matching, and determines the character in the input image. Based on the degree of coincidence determined at each relative position by the information determining unit that acquires character region information indicating each region of the image and the matching determining unit, the character region information acquired by the information acquiring unit is corrected. And an encoding unit that encodes the character region information of the character image corrected by the region correction unit and the identification information of the image pattern corresponding to the character image in association with each other. To do.

[Image dictionary creation method]
The image dictionary creation method according to the present invention provides a typical character image as a template image, determines the degree of matching between the character image included in the input image and the provided template image, and the determined match Based on the degree, the provided template image is determined as a typical image pattern in the input image, and identification information for identifying each image pattern is given to the determined image pattern.

[program]
In addition, the program according to the present invention provides a step of providing a typical character image as a template image in an image dictionary creation device including a computer, and a degree of coincidence between the character image included in the input image and the provided template image. Determining the provided template image as a typical image pattern in the input image based on the determined degree of matching, and determining each image pattern for the determined image pattern A step of providing identification information for identifying the image dictionary.

  According to the encoding device of the present invention, a high compression rate can be realized while maintaining the readability of characters.

[First Embodiment]
First, in order to help understanding of the present invention, its background and outline will be described.
For example, instead of encoding the character image itself included in the input image, the image processing apparatus 2 realizes a high compression rate by encoding the identification information of each character image and its appearance position. Can do.
FIG. 1A illustrates an encoding method on the assumption that a common font database exists, and FIG. 1B illustrates an encoding method on the assumption that an image dictionary is attached. .
As shown in FIG. 1A, when there is a common font database that stores character images in association with identification information (character code and font type) on both the encoding side and the decoding side, The image processing apparatus on the encoding side encodes the character image identification information (character code, font type, etc.) and the appearance position of the character image, thereby compressing the image data at a high compression rate and decoding the image data. To the image processing apparatus. In this case, the image processing apparatus on the decoding side decodes the received code data (character code and font type), decodes the character code and font type, and the font registered in the font database. A character image is generated based on the image.
However, in the encoding method based on the presence of the font database, it is necessary to provide a font database on each of the encoding side and the decoding side, and the storage area is compressed by the font database. Further, when the encoding-side font database is updated, it is necessary to update the decoding-side font database so as to have the same contents as the encoding-side.

Therefore, as shown in FIG. 1B, the image processing apparatus 2 according to the present embodiment registers, on the decoding side, image patterns that exist typi- cally in the input image in association with indexes (image dictionary formation). Then, the image pattern included in the input image is replaced with the corresponding index and its appearance position and encoded. The encoding side transmits the image dictionary in which the image pattern and the index are associated with each other, and the encoded index and the appearance position to the decoding side. The decoding side decodes the index and the appearance position, selects an image pattern corresponding to the decoded index from the image dictionary, and arranges it at the decoded appearance position.
As described above, the image processing apparatus 2 can achieve a high compression rate without assuming a common database by creating and transmitting / receiving an image dictionary according to an input image. Further, it is not necessary to synchronize the font database on the encoding side and the decoding side. In order to reduce the code amount, it is desirable that the image dictionary is also encoded.

FIG. 2A illustrates an image dictionary, and FIG. 2B illustrates an image pattern.
As illustrated in FIG. 2A, the image dictionary includes a plurality of image patterns included in the input image and an index assigned to identify the image pattern. The image pattern is partial image data included in the input image. In this example, the image pattern is a typical pattern (binary data) that appears more than a predetermined number (multiple times) in the input image (binary). The index is, for example, identification information generated individually for each input image, and may be a serial number assigned to the image pattern in the order in which the image pattern is extracted from the input image.

The next problem is how to extract an image pattern from an input image and register it as an image dictionary. This is because the code amount of the input image varies depending on the size and appearance frequency of the extracted image pattern. For example, as illustrated in FIG. 2B, there are a case where an image pattern is extracted in units of character images and a case where an image pattern is extracted in units smaller than a character image.
When an image pattern is extracted in a smaller unit than a character image, the appearance frequency of each image pattern is often high (for example, the vertical bar portion of “1” indicates “mountain” and “river”. The number of image patterns to be registered in the image dictionary increases, and the data amount of the image dictionary increases.
On the other hand, when an image pattern is extracted in units of character images, a large number of characters of the same font type and the same font size appear in the same language in the same document. Can expect high appearance frequency.
In addition, when aiming for a high compression rate while allowing a certain degree of irreversibility, the image processing apparatus on the encoding side replaces not only the partial image that is the same as the image pattern but also the partial image that is similar to the image pattern with an index. To encode. In this case, if each of the constituent parts of the character image is replaced with a similar image pattern, the entire character image may be decoded into a completely different one and lose readability. However, when an image pattern is extracted in units of character images, the overall shape of the character image is replaced with a similar image pattern (for example, the number “1” and the alphabet “I”). Some readability is maintained.
Therefore, the image processing apparatus 2 in this embodiment determines an image pattern for each character image and registers it in the image dictionary.
Further, the image processing apparatus 2 according to the present embodiment creates an image dictionary by applying a font image stored in a font database as an image pattern template. This is because the input image to be encoded includes many character images composed of font images, and even if a certain handwritten character or the like is replaced with a font image, the readability is not affected.
Furthermore, the image processing apparatus 2 in the present embodiment creates an image pattern based on the font image. This eliminates image quality degradation due to repeated copying and improves readability. Note that the image processing apparatus 2 of the present embodiment only uses a font image in the encoding process, and does not need to use a font database at the time of decoding.

[Hardware configuration]
Next, the hardware configuration of the image processing apparatus 2 will be described.
FIG. 3 is a diagram illustrating a hardware configuration of the image processing apparatus 2 to which the image dictionary creating method according to the present invention is applied, centering on the control apparatus 20.
As illustrated in FIG. 3, the image processing apparatus 2 includes a control device 20 including a CPU 202 and a memory 204, a communication device 22, a recording device 24 such as an HDD / CD device, an LCD display device or a CRT display device, and a keyboard. A user interface device (UI device) 26 including a touch panel and the like is included.
The image processing apparatus 2 is, for example, a general-purpose computer in which an encoding program 5 (described later) is installed as a part of a printer driver. The image processing apparatus 2 acquires image data via the communication device 22 or the recording device 24 and the like. The image data is encoded and transmitted to the printer apparatus 10. Further, the image processing device 2 acquires image data optically read by the scanner function of the printer device 10 and encodes the acquired image data.

[Encoding program]
FIG. 4 is a diagram illustrating a functional configuration of the encoding program 5 which is executed by the control device 20 (FIG. 3) and implements the image dictionary creation method according to the present invention.
As illustrated in FIG. 4, the encoding program 5 includes an image input unit 40, an image dictionary creation unit 50, and an encoding unit 60.
In the encoding program 5, the image input unit 40 is an image data read by the scanner function of the printer device 10, or a PDL (Page Discription Language) format image acquired via the communication device 22 or the recording device 24. Data is acquired, and the acquired image data is converted into raster data and output to the image dictionary creation unit 50. The image input unit 40 includes a character recognition unit 410 that recognizes a character image from optically read image data and the like, and a PDL decomposer 420 that interprets PDL format image data and generates raster data.
The character recognition unit 410 recognizes characters included in the input image data (hereinafter referred to as input image), and character recognition information of the recognized characters and character area information of the recognized characters are obtained as a result of character recognition processing. Is output to the image dictionary creation unit 50. Here, the character identification information is information for identifying a character, and is, for example, a versatile character code (such as an ASCII code or a shift JIS code) or a combination of a character code and a font type. The character area information is information indicating the area of the character image in the input image, and is, for example, character layout information including the position, size, range, or combination of the character images.
The PDL decomposer 420 interprets the image data in the PDL format, generates rasterized image data (raster data), and character identification information and character area information of the character image in the generated image data together with the generated image data Is output to the image dictionary creation unit 50.

  The image dictionary creation unit 50 creates an image dictionary used for encoding the input image based on the input image input from the image input unit 40 and a font image stored in advance, and the created image The dictionary and the input image are output to the encoding unit 60. More specifically, the image dictionary creation unit 50 selects a font image corresponding to the character image as an image pattern based on the character identification information and the character region information input from the character recognition unit 410 or the PDL decomposer 420, An index is assigned to the selected font image, and an image dictionary is output to the encoding unit 60.

  The encoding unit 60 encodes an input image based on the image dictionary input from the image dictionary creation unit 50, and stores the encoded input image and the image dictionary in the recording device 24 (FIG. 3) or the printer device 10 (FIG. 3) and so on. More specifically, the encoding unit 60 compares the image pattern registered in the image dictionary with the partial image included in the input image, and converts the partial image data that matches one of the image patterns to this data. The index corresponding to the image pattern and the position information of this partial image are replaced. Here, the term “match” includes not only a complete match but also a case where a partial match is made within a predetermined allowable range. Furthermore, the encoding unit 60 may encode the index and position information replaced with the partial image, the image dictionary, and the like by entropy encoding (Huffman encoding, arithmetic encoding, LZ encoding, or the like).

FIG. 5 is a diagram for explaining the function of the first image dictionary creation unit 50 in more detail.
As shown in FIG. 5, the image dictionary creation unit 50 includes a storage unit 500, a character image extraction unit 510, a font provision unit 520, a font database 525 (hereinafter, font DB 525), a match determination unit 530, a dictionary determination unit 540, a position A correction unit 550, an index assignment unit 560, and a registration control unit 570 are included. The storage unit 500 controls the memory 204 (FIG. 3) and the recording device 24 (FIG. 3) to store the input image, character identification information, and character region information input from the image input unit 40 (FIG. 4). Hereinafter, a character code will be described as a specific example of character identification information, and character position information will be described as a specific example of character region information.

  The character image extraction unit 510 cuts out a character image from the input image based on the character position information. That is, the character image extraction unit 510 extracts an area indicated by the character area information as a character image from the input image. The extracted character image is an area determined by the character recognition unit 410 to be a character image. The character recognition unit 410 or the PDL decomposer 420 may output the character image to the image dictionary creation unit 50 in a state where the character image is cut out from the input image.

  The font providing unit 520 provides the font image stored in the font DB 525 as a template image to the matching determination unit 530, the dictionary determination unit 540, and the like. Specifically, the font DB 525 stores a plurality of font images in which characters corresponding to the character code are expressed in a plurality of fonts and sizes in association with the character code. The font providing unit 520 selects at least one font image from the font images stored in the font DB 525 based on the character code input from the character recognition unit 410, and uses the selected font image as a template image. Store in the storage unit 500.

The matching determination unit 530 compares the character image cut out from the input image with the font image provided from the font providing unit 520 to determine the degree of matching. Here, the degree of coincidence is information indicating the degree to which a plurality of images coincide with each other. For example, when binary images are compared, the number of pixels that overlap each other when the two images are overlaid (hereinafter referred to as the number of pixels). , The number of matching pixels), the matching pixel ratio obtained by normalizing the number of matching pixels (for example, the number of matching pixels divided by the total number of pixels), or the pixel distribution (histogram) when a plurality of images are superimposed is there.
Further, the coincidence determination unit 530 compares the character image and the font image at a plurality of relative positions to determine the degree of coincidence. That is, the coincidence determination unit 530 performs comparison while changing (shifting) the relative position between the character image and the font image in order to calculate the maximum degree of coincidence.
For example, the coincidence determination unit 530 calculates a coincidence pixel rate while shifting a character image cut out from the input image and a font image selected based on the character code of the character image, and the maximum value of the coincidence pixel rate The shift vector when the maximum value is reached is output to the storage unit 500.
Similarly, the coincidence determination unit 530 compares the character image cut out from the input image with the image pattern registered in the image dictionary to determine the degree of coincidence.

  The dictionary determining unit 540 determines an image pattern to be registered in the image dictionary based on the degree of matching determined by the matching determining unit 530 and the font image provided from the font providing unit 520. Specifically, the dictionary determination unit 540 selects, from among font images selected from the font providing unit 520 based on the character code, a font image having the highest degree of matching determined by the matching determination unit 530 as an image pattern. To do.

The position correction unit 550 corrects the position information of the character image based on the shift vector output from the match determination unit 530. That is, the position correction unit 550 corrects the position information input from the image input unit 40 so that the degree of matching between the input image and the font image is maximized.
The index assigning unit 560 assigns an index for identifying these image patterns to the image pattern determined based on the font image, and outputs the assigned index and the image pattern to the storage unit 500 in association with each other. .

  The registration control unit 570 determines whether or not to newly register an image pattern based on the degree of matching between the character image extracted by the character image extraction unit 510 and the image pattern, and controls new registration in the image dictionary. To do. Specifically, the registration control unit 570 uses the font image corresponding to the input character image as an image pattern based on the degree of matching between the sequentially input character image and the image pattern registered in the image dictionary. Judge whether or not to register. For example, the registration control unit 570 displays the character when the degree of coincidence between the registered image pattern and the input character image is less than or equal to the reference (that is, when a matching image pattern is not registered). When a font image provided for an image is allowed to be registered in the image dictionary as an image pattern, and there is an image pattern whose matching degree is greater than the reference (that is, when a matching image pattern exists) The font image corresponding to the character image is prohibited from being registered in the image dictionary as an image pattern.

FIG. 6 is a diagram for explaining the function of the encoding unit 60 in more detail.
As shown in FIG. 6, the encoding unit 60 includes a pattern determination unit 610, a position information encoding unit 620, an index encoding unit 630, an image encoding unit 640, a dictionary encoding unit 650, a selection unit 660, and a code output unit. 670.
The pattern determination unit 610 compares each image pattern registered in the image dictionary with a partial image included in the input image, and determines an image pattern (same or similar image pattern) corresponding to the partial image. Specifically, the pattern determination unit 610 superimposes the partial image (corrected by the position correction unit 550) cut out in units of character images from the input image and the image pattern, and matches the determination unit 530 ( The degree of coincidence is calculated by the same method as in FIG. 5), and it is determined whether or not it corresponds based on whether or not the calculated degree of coincidence is equal to or greater than a reference value.
When the corresponding image pattern is found, the pattern determination unit 610 outputs the position information of the partial image to the position information encoding unit 620, and the index of the image pattern is output to the index encoding unit 630. If the corresponding image pattern is not found, the partial image is output to the image encoding unit 640.
Note that the pattern determination unit 610 in the present embodiment uses the index of the image pattern (that is, the font image) that matches the input character image, and the position information of the character image (corrected by the position correction unit 550). When the font image corresponding to the input character image exists in the font DB 525 for acquisition from the image dictionary creation unit 50, the acquired index and position information are respectively converted into the index encoding unit 630 and the position information encoding unit. If there is no font image corresponding to the input character image, the character image itself is output to the image encoding unit 640.

The position information encoding unit 620 encodes the position information input from the pattern determination unit 610 (that is, position information of the partial image (character image) corrected by the position correction unit 550), and outputs the encoded position information to the selection unit 660. To do. For example, the position information encoding unit 620 encodes position information by applying LZ encoding or arithmetic encoding.
The index encoding unit 630 encodes the index input from the pattern determination unit 610 and outputs the encoded index to the selection unit 660. For example, the index encoding unit 630 assigns a code having a different code length to each index according to the appearance frequency of the index.
The image encoding unit 640 encodes the partial image input from the pattern determination unit 610 by applying an encoding method suitable for the image, and outputs the encoded partial image to the selection unit 660.
The dictionary encoding unit 650 encodes the image dictionary (the image pattern and the index are associated with each other) input from the image dictionary creation unit 50 (FIGS. 4 and 5), and the code output unit 670 Output.

When the pattern determination unit 610 finds an image pattern corresponding to the partial image, the selection unit 660 receives the position information code data input from the position information encoding unit 620 and the index encoding unit 630. The index code data is output to the code output unit 670 in association with each other, and is encoded by the image encoding unit 640 when the pattern determination unit 610 finds no image pattern corresponding to the partial image. The code data of the partial image is output to the code output unit 670.
The code output unit 670 receives the code data (position information, index, and partial image code data) input from the selection unit 660 and the code data (image dictionary code data) input from the dictionary encoding unit 650. The data are output to the printer device 10 (FIG. 3), the recording device 24 (FIG. 3), or the communication device 22 (FIG. 3) in association with each other.

[Encoding operation]
Next, the overall operation of the encoding process by the image processing apparatus 2 will be described.
FIG. 7 is a flowchart showing the operation (S10) of the encoding program 5 in the first embodiment.
As shown in FIG. 7, in S100, when image data (binary) is input from the printer 10 (FIG. 3), the image input unit 40 (FIG. 4) receives the input image data (input image). It outputs to the image dictionary creation part 50. Further, the character recognition unit 410 (FIG. 4) of the image input unit 40 performs character recognition processing on the input image, and determines the character code and position information of the character image included in the input image one character at a time. The obtained character code and position information are output to the image dictionary creation unit 50.

In step 102 (S102), the storage unit 500 (FIG. 5) of the image dictionary creation unit 50 stores the input image (binary), character code, and position information input from the image input unit 40 in the memory 204 (FIG. 3). Remember.
The character image extraction unit 510 (FIG. 5) cuts out character images for each character from the input image based on the input position information and outputs them to the storage unit 500.
In step 104 (S104), the registration control unit 570 reads an image pattern and an index of the image pattern from the image dictionary based on the character code of the character image cut out in order from the input image. Specifically, the storage unit 500 stores an already determined image pattern, an index assigned to the image pattern, and a character code of a character image corresponding to the image pattern as an image dictionary. doing. The registration control unit 570 acquires the character code of the character image newly cut out by the character image extraction unit 510 from the character recognition unit 410, and reads an image pattern and an index corresponding to the acquired character code from the image dictionary.

In step 106 (S106), the coincidence determination unit 530 compares the character images sequentially extracted by the character image extraction unit 510 with the image patterns registered in the image dictionary, and determines the degree of coincidence at a plurality of relative positions. To do. Specifically, the coincidence determination unit 530 calculates the number K of coincident pixels of black pixels while shifting each image pattern registered in the image dictionary and the extracted character image.
The coincidence pixel number K is a position vector indicating a relative position in the image, x is a black pixel distribution of the image pattern, S (x), a character image number cut out in order is i (1 to N), When the distribution of black pixels is P (i, x) and the shift vector of the character image i is vi, it is calculated by the following formula.
(Number of matched pixels K) = Σ {S (x) * P (i, x−vi)}
Note that “Σ” indicates the total sum for the variable x.
Next, the coincidence determination unit 530 normalizes the calculated coincidence pixel number K to calculate a coincidence pixel rate K ′.
The coincidence pixel rate K ′ is calculated by the following equation, where M is the number of pixels constituting the character image.
(Matched pixel rate K ′) = K / M

In step 108 (S108), the registration control unit 570 converts the image pattern determined by the match determination unit 530 into a newly cut out character image based on the degree of matching between the image pattern and the character image (newly cut out). Based on this, it is determined whether or not to register an image pattern. Specifically, when the determined matching pixel rate K ′ is greater than or equal to the reference value, the registration control unit 570 includes the index of the image pattern having the largest matching pixel rate K ′ and the position information of the extracted character image. Are output to the encoding unit 60, and the process proceeds to S122. If the determined matching pixel ratio K ′ is smaller than the reference value, the process proceeds to S110.
That is, the registration control unit 570 prohibits registration of the image pattern based on the character image and encodes the character image based on the registered image pattern when the matching degree is equal to or higher than the reference. If the degree of coincidence is smaller than the reference, the dictionary determining unit 540 is instructed to newly register a font image corresponding to the character image as an image pattern in the image dictionary.

  In step 110 (S110), the font providing unit 520 reads a font image corresponding to the extracted character image from the font DB 525 based on the character code input from the character recognition unit 410, and reads the read font image ( 1 or more) is output to the storage unit 500.

In step 112 (S112), the coincidence determination unit 530 compares the font image provided by the font providing unit 520 and the clipped character image while relatively moving them to determine the degree of coincidence. Specifically, the coincidence determination unit 530 calculates the number K of coincidence pixels between the provided one or more font images T (x) and the extracted character image P (i, x-vi).
(Number of matched pixels K) = Σ {T (x) * P (i, x−vi)}
Next, the coincidence determination unit 530 calculates a coincidence pixel rate K ′ based on the coincidence pixel number K. (Matched pixel rate K ′) = K / M

  In step 114 (S114), the dictionary determination unit 540 determines whether or not to register the font image as an image pattern in the image dictionary based on the degree of matching between the font image and the character image determined by the match determination unit 530. To do. Specifically, when the matching pixel rate K ′ between the font image and the character image is equal to or greater than a predetermined threshold, the dictionary determining unit 540 proceeds to the process of S116, and the matching pixel rate K ′ If it is smaller, the process proceeds to S126.

  In step 116 (S116), the dictionary determining unit 540 selects a font image to be registered as an image pattern based on the degree of matching determined for each font image. Specifically, the dictionary determining unit 540 compares the matching pixel rate K ′ calculated for the cut character image and each font image, and determines the font image having the maximum matching pixel rate K ′ as an image pattern. Choose as.

In step 118 (S118), the index assigning unit 560 assigns identification information (index) for uniquely identifying the image pattern to the image pattern selected by the dictionary determining unit 540. Then, the index adding unit 560 stores the added index in the storage unit 500 as an image dictionary in association with the image pattern. The assigned index uniquely identifies each image pattern for at least all input images input as encoding targets.
The assigned index is sequentially output to the encoding unit 60 as encoding target data.

In step 120 (S120), the position correction unit 550 receives an input from the character recognition unit 410 based on the degree of matching (the number of matching pixels K or the matching pixel rate K ′) calculated at the plurality of relative positions by the matching determination unit 530. A correction vector for the obtained position information is determined. Specifically, in S112, the position correction unit 550 outputs the shift vector vi (input from the character recognition unit 410) when the number of coincident pixels K calculated for the clipped character image and font image is the maximum. A two-dimensional vector obtained by displacing the character image with reference to the position information) is set as a correction vector.
That is, when the image pattern corresponding to the clipped character image is not registered in the image dictionary, the image dictionary creation unit 50 uses the font image registered in the image dictionary as a reference and the font image and the character image are the most. Correct the cutting position so that they match.
Note that the position information of the character image corrected by the correction vector is sequentially output to the encoding unit 60 as encoding target data.

In step 122 (S122), the position correction unit 550 determines the correction vector of the position information input from the character recognition unit 410 based on the degree of matching determined by the match determination unit 530 for the image pattern and the character image. . Specifically, in S106, the position correction unit 550 sets the shift vector vi when the number of matching pixels K calculated for the extracted character image and image pattern is the maximum as the correction vector.
That is, when the image pattern corresponding to the clipped character image is already registered in the image dictionary, the image dictionary creation unit 50 cuts out the clipped position so that the registered image pattern and this character image most closely match. Correct.
Note that the position information of the character image corrected by the correction vector is sequentially output to the encoding unit 60 as encoding target data.

  In step 124 (S124), the encoding unit 60 encodes the index and character image position information (corrected by the position correction unit 550) input as encoding target data from the image dictionary creation unit 50, and The converted index and position information (code data) are sequentially output to the printer 10 or the recording device 24 or the like. Specifically, the position information encoding unit 620 encodes the position information input from the image dictionary creation unit 50, and the index encoding unit 630 encodes and selects the index input from the image dictionary creation unit 50. The unit 660 and the code output unit 670 output the encoded position information and index in association with each other.

In step 126 (S126), the encoding unit 60 encodes the extracted character image data as it is, and outputs the encoded image data (code data) to the printer device 10 or the recording device 24 or the like. Specifically, the image encoding unit 640 encodes the extracted character image data as it is, and the selection unit 660 and the code output unit 670 output the encoded character image data.
That is, the encoding unit 60 encodes the extracted character image itself when the extracted character image and the font image provided for the character image do not match at least the reference.

  In step 128 (S128), the encoding program 5 determines whether or not the encoding process has been completed for all character images included in the input image. If all the character images have been completed, the process of S130 is performed. In other cases, the process returns to S102 and the next character image is cut out, and the processes from S104 to S128 are repeated.

In step 130 (S130), the encoding unit 60 encodes the image pattern and index registered as an image dictionary. Specifically, the dictionary encoding unit 650 encodes the image pattern and the index input from the image dictionary creation unit 50 as an image dictionary in association with each other.
In step 132 (S132), the encoding unit 60 associates the encoded image dictionary with the code data of the input image and outputs it to the printer device 10 or the recording device 24 or the like. Specifically, the code output unit 670 outputs the image pattern and index (that is, the image dictionary) encoded by the dictionary encoding unit 650 in association with the code data of the input image.

As described above, the image processing apparatus 2 according to the first embodiment compares the character image included in the input image with the font image, and registers the font image in the image dictionary based on the degree of matching. This makes it possible to efficiently select a typical image pattern included in the input image.
The image processing apparatus 2 selects a font image based on the character code acquired as a result of the character recognition process. As a result, the font images to be compared with the cut-out character image are narrowed down, and the font image to be registered in the image dictionary can be easily selected. Note that the image processing apparatus 2 selects a font image according to the font type or the image size or the like when the font type or the image size (for example, the font size) of the character image is acquired as a result of the character recognition process. Alternatively, the font image may be selected based on the character identification information generated by the PDL decomposer 420.
In addition, the image processing apparatus 2 according to the present embodiment registers a font image closest to the character image in the image dictionary as an image pattern. As a result, the smearing of the original and the image quality deterioration due to repeated copying are eliminated, and the readability is improved.
Further, the image processing apparatus 2 in the present embodiment corrects the cutout position of the character image with reference to the font image. Thereby, the position of each character image is more accurately reproduced in the decoded image. In particular, it is suitable when the position of the center of gravity differs between the character cut-out area and the character to be cut out.

[Second Embodiment]
Next, a second embodiment will be described.
In the first embodiment, the image dictionary creation unit 50 sequentially cuts out character images from the input image, and sequentially creates an image dictionary based on the cut out character images. The image dictionary creation unit 52 creates an image dictionary based on the entire input image such as one page or one document.
In addition, the image dictionary creation unit 52 according to the present embodiment creates an image pattern to be registered in the image dictionary based on a plurality of font images having similar forms, so that the image dictionary creation unit 52 is highly compatible with the variety of font images. An image dictionary that realizes encoding efficiency is created.

FIG. 8 is a diagram for explaining the relationship between a character code and a font image.
As illustrated in FIG. 8A, each character code is associated with a plurality of font images having different shapes. For example, a plurality of font images having different shapes can be provided based on a character code corresponding to the numeral “1”. In the first embodiment, when a character image that matches each of these font images exists in the input image, each font image is registered as an image pattern.
However, these font image differences do not significantly affect the readability of characters. Also, many of the same fonts are applied in the same document, and other fonts are rarely applied in headings and the like. Even for a font image having a low appearance frequency, if an image pattern is registered in the image dictionary, improvement in coding efficiency is hindered.
Further, as illustrated in FIG. 8B, due to a mistake in character recognition processing, characters having similar shapes (for example, the number “1” and the capital letter “I”) can be identified as the same character code. In this case, since the shape is at least approximate, the influence on the readability is not so great, but there are cases where it is desired to reproduce these so that they can be distinguished.

  Thus, the image processing apparatus 2 according to the second embodiment creates an image pattern to be registered in the image dictionary by integrating a plurality of font images similar to each other in accordance with the appearance frequency in the input image. . For example, the image processing apparatus 2 integrates font images corresponding to the same character code in accordance with the appearance frequency of the character image corresponding to each font image, and creates an image pattern that frequently appears in the input image. There may be a plurality of image patterns created from font images having the same character code.

FIG. 9 is a diagram illustrating the functional configuration of the second image dictionary creation unit 52. Of the components shown in the figure, those substantially the same as those shown in FIG. 5 are given the same reference numerals.
As shown in FIG. 9, the image dictionary creation unit 52 in the second embodiment has a configuration in which the registration control unit 570 is replaced with a character classification unit 580 in the first image dictionary creation unit 50 shown in FIG. .
The character classifying unit 580 classifies the character images cut out from the input image into a plurality of character image groups based on the character code. For example, the character classification unit 520 classifies character images having the same character code into the same character image group.
In this case, the dictionary determining unit 540 responds to the appearance frequency of each font image based on the font images corresponding to the plurality of character images having the same character code (the font image most matching each of the plurality of character images). Create an image pattern. Note that the character classification unit 580 may classify the font images corresponding to each of the extracted character images (the font image that most closely matches each character image) according to the character code of the extracted character image.

FIG. 10 is a flowchart showing the second operation (S20) of the encoding program 5. In this flowchart, a case where binary image data optically read by the scanner function of the printer apparatus 10 is input will be described as a specific example. Also, among the processes shown in the figure, the same reference numerals are assigned to the processes that are substantially the same as those shown in FIG.
As shown in FIG. 10, in the processing from S100 to S112, when image data (binary) is input from the printer 10 (FIG. 3), the image input unit 40 converts the input image data (input image). Character recognition processing is performed on the image, and the result of the character recognition processing (character code and position information) and the input image are output to the image dictionary creation unit 50.
The image dictionary creation unit 52 cuts out character images one character at a time from the input image based on the input position information, and reads out one or more font images from the font DB 525 based on the character code of the cut out character image. The matching pixel ratio K ′ is calculated while relatively moving the read font image (one or more) and the extracted character image.

In step 202 (S202), the dictionary determination unit 540 determines from the font images read from the font DB 525 based on the matching pixel rate K ′ between the font image and the character image calculated by the matching determination unit 530. Select the font image that most closely matches the extracted character image.
In step 204 (S204), the image dictionary creation unit 50 determines whether or not all the character images included in the input image have been cut out. If all the character images have been cut out, the process proceeds to S206. In other cases, the process returns to S102, the next character image is cut out, and the processes from S110 to S204 are repeated.

In step 206 (S206), the character classification unit 580 classifies the font images selected for each of the cut out character images (the font image that most closely matches each character image) based on the character codes of these character images. .
In step 210 (S210), the dictionary determination unit 540 creates an image pattern to be registered in the image dictionary based on the font images classified by the character classification unit 580.

In step 230 (S230), the index assigning unit 560 assigns an index to the image pattern created by the dictionary determining unit 540, and stores the assigned index in the storage unit 500 in association with the image pattern. The assigned index uniquely identifies each image pattern in at least all input images input as encoding targets.
When image pattern creation and index assignment have been completed for all input images input as encoding targets, these image patterns and indexes are output to the encoding unit 60 as an image dictionary.

In step 240 (S240), the encoding unit 60 compares the image pattern registered in the image dictionary with the partial image included in the input image, and an image pattern that matches the partial image exists in the image dictionary. In addition, this partial image is encoded by replacing it with the index and position information, and when the image pattern matching the partial image does not exist in the image dictionary, the partial image is encoded as it is. The encoding unit 60 encodes the image dictionary.
The encoding unit 60 associates the index, the position information, the partial image, and the code data of the image dictionary with each other and outputs them to the printer device 10 or the like.

FIG. 11 is a flowchart illustrating the image pattern creation process (S210) in the second embodiment in more detail.
As shown in FIG. 11, in step 212 (S212), the coincidence determination unit 530 compares the font images classified for each character code with each other, and determines the degree of coincidence at a plurality of relative positions. Specifically, the coincidence determination unit 530 creates a pixel distribution (histogram) of black pixels in a plurality of classified font images (hereinafter referred to as font image group), and creates the pixel distribution and the font image group. The number of matching pixels of black pixels is calculated while shifting the contained font image from each other. The pixel distribution is a histogram obtained by sequentially adding pixel values for each region at a relative position where the number of matching pixels is maximum for the black pixels of the font image belonging to the font image group.
That is, the pixel distribution of the font image group is Q (x), the black pixel distribution of each font image is P (i, x), the position vector is x, and each font image belonging to the font image group is i (1 to N: N). Is the number of font images belonging to the font image group), and when the shift vector of the font image i is vi, the coincidence determination unit 530 calculates the number of coincidence pixels by the following formula.
(Number of matched pixels K) = Σ {Q (x) * P (i, x−vi)}
When i = 1,
Q (x) = P (1, x)
And
If i> 1,
Q (x) = P (1, x) + P (2, x−v2) +... + P (i−1, x−v (i−1))
It becomes.

In step 214 (S214), the dictionary determining unit 540 performs threshold processing for removing the number of distributions equal to or less than the threshold for the degree of matching (pixel distribution) calculated by the matching determining unit 530. Specifically, the dictionary determining unit 540 normalizes Q (x) calculated by the match determination unit 530 to calculate Q ′ (x), and performs threshold processing on the calculated Q ′ (x). Do. That is, the dictionary determination unit 540 calculates the distribution probability Q ′ (x) by the following formula.
Q ′ (x) = Q (x) / N
Next, the coincidence determination unit 530 calculates Q ″ (x) by removing a portion where the distribution probability Q ′ (x) is smaller than the reference value according to the following conditional expression.
When Q ′ (x)> threshold A, Q ″ (x) = 1
In other cases, Q ″ (x) = 0

In step 216 (S216), the dictionary determining unit 540 determines whether or not an area where the number of distributions is not 0 (black pixel area) is wider than the reference with respect to the pixel distribution after the threshold processing. The process proceeds to S218, and if it is narrower than the reference, the image pattern creation process (S210) is terminated without registering the image pattern for this font image group. Specifically, the dictionary determination unit 540 determines whether or not the number of pixels for which Q ″ (x) is 1 is greater than or equal to a reference value. If the number is greater than or equal to the reference value, registration of an image pattern is performed. If it is smaller than the reference value, the image pattern is not registered.
That is, the image processing apparatus 2 encodes a character image itself without registering an image pattern for a font image (that is, a character image) with a small number of appearances. This eliminates registration of unnecessary image patterns from the viewpoint of encoding efficiency.

In step 218 (S218), the dictionary determining unit 540 determines the sum combination pattern Q ″ (x) based on the pixel distribution. Specifically, the dictionary determining unit 540 performs Q ″ ( Let x) be the union pattern.
In step 220 (S220), the dictionary determining unit 540 extracts a common part (product set) between the sum combination pattern Q ″ (x) and the font images belonging to the font image group as a common pattern.
In step 222 (S222), the dictionary determining unit 540 determines the extracted common pattern as an image pattern. Note that the dictionary determining unit 540 does not register all the common patterns extracted based on the respective font images in the image dictionary, but excludes overlapping common patterns and registers them in the image dictionary. Further, the dictionary determination unit 540 may also exclude a plurality of common patterns extracted based on each character image as overlapping even when the difference in the black pixel area is equal to or less than a threshold value. The dictionary determination unit 540 may register the sum combination pattern Q ″ itself as an image pattern in the image dictionary.

In step 224 (S224), the coincidence determination unit 530 outputs a plurality of character images that have been cut out and image patterns corresponding to the character images (created from font images classified by character codes of the character images). The number of coincidence pixels K is calculated by comparison at the relative position.
The position correction unit 550 sets the shift vector vi when the number of matching pixels K calculated by the matching determination unit 530 is maximum as a correction vector.

FIG. 12 is a diagram schematically illustrating the image pattern creation process (S210) in the second embodiment.
As shown in FIG. 12, the black pixel distribution probability Q ′ (x) of the font image classified by the character code indicates a different numerical value depending on the pixel position x. This is because font images having different shapes are mixed with different appearance frequencies in the font image group. The distribution probability Q ′ (x) is subjected to threshold processing using the threshold B, and thus a portion with a low distribution probability is removed. Thereby, it is possible to eliminate only the differential shape (difference from the font image with high appearance frequency), the noise portion, and the like of the font characters (input image) with low appearance frequency, and make only a typical shape as an image pattern.
Subsequently, the dictionary determination unit 540 generates a sum coupling pattern Q ″ (x) in which a portion equal to or greater than the threshold B is 1 and a portion smaller than the threshold B is 0, and the generated sum coupling pattern Q ″ (x ) And each font image belonging to the font image group is extracted as a common pattern. That is, the dictionary determining unit 540 performs a product operation on the sum coupling pattern Q ″ (x) and the pixel distribution P (i, x−vi) of each font image. Thereby, a plurality of font images exist in the font image group. Image pattern # 1 and image pattern # 2 of a typical font image (a font image having a high appearance frequency) are extracted.

FIG. 13 is a flowchart for explaining in more detail the encoding process (S240) in the second embodiment. In this flowchart, a case where the encoding process is performed based on the image pattern created in FIG. 11 will be described as a specific example.
As shown in FIG. 13, in step 242 (S242), the pattern determination unit 610 uses a partial image (that is, a character image) cut out from the input image based on the position information corrected by the correction vector, and an image dictionary. The number of matching pixels is calculated by comparing with the registered image pattern. Note that the pattern determination unit 610 may acquire the number of matching pixels from the matching determination unit 530.
In step 244 (S244), the pattern determination unit 610 determines whether there is a matching image pattern. Specifically, the pattern determination unit 610 determines whether or not the number of matching pixels calculated for each image pattern is within an allowable range (for example, 90% or more with respect to all pixels of the partial image), If it is within the allowable range, the process proceeds to S246, and if it is outside the allowable range, the process proceeds to S250.

In step 246 (S246), the pattern determination unit 610 reads the index of the image pattern having the largest number of matching pixels from the image dictionary among the image patterns having the matching pixel number within the allowable range, and reads the read index into the index. This is output to the encoding unit 630 and the position information of the character image (corrected by the position correction unit 550) is output to the position information encoding unit 620.
The index encoding unit 630 encodes the index input from the pattern determination unit 610 and outputs the index code data to the selection unit 660.

In step 248 (S248), the position information encoding unit 620 encodes the position information input from the pattern determination unit 610 and outputs code data of the position information to the selection unit 660.
The selection unit 660 associates the code data of the index input from the index encoding unit 630 and the code data of the position information input from the position information encoding unit 620 with each other and outputs them to the code output unit 670. . That is, the selection unit 660 outputs the index and the position information to the code output unit 670 so as to be associated with each other for each partial image.

In step 250 (S250), the pattern determination unit 610 outputs this partial image (that is, a character image for which the corresponding image pattern does not exist in the image dictionary) to the image encoding unit 640.
The image encoding unit 640 encodes the partial image image data input from the pattern determination unit 610 and outputs the partial image code data to the selection unit 660.
The selection unit 660 outputs the code data of the partial image input from the image encoding unit 640 to the code output unit 670.

In step 252 (S252), the pattern determination unit 610 determines whether or not the encoding process has been completed for all partial images. If there is an unencoded partial image, the process returns to the process of S242. When the next partial image is encoded and all the partial images have been encoded, the process proceeds to S254.
In step 254 (S254), the dictionary encoding unit 650 encodes the image dictionary (the image pattern and the index associated with each other) input from the image dictionary creation unit 50, and outputs the code data of the image dictionary as code output. Output to the unit 670.

As described above, the image processing apparatus 2 according to the second embodiment has a font image corresponding to each character image included in the input image (the font image is read out for each character image. A plurality of types), and a plurality of typical image patterns are determined based on the classified font images and the appearance frequency of these font images. Can do. As a result, if the appearance frequency is high, font images having different fonts or sizes are also registered in the image dictionary as image patterns, and font images with a low appearance frequency are not registered in the image dictionary. Encoding efficiency can be improved.
In the present embodiment, the image processing apparatus 2 classifies the character images into character image groups based on the character codes. However, the image processing apparatus 2 may classify the character images using character codes and character image sizes.

[Modification]
Next, a modification of the second embodiment will be described.
FIG. 14 is a diagram illustrating common shapes and differential shapes of a plurality of character images that can be determined as the same character code.
As illustrated in FIG. 14A, a plurality of font images having different shapes (that is, different font types) are associated with the same character code. Thus, the font image group corresponding to the character code has the same shape in most cases. Hereinafter, a partial image shared by a plurality of font images belonging to a font image group is referred to as a common shape, and a partial image that is a difference between each font image and the common shape is referred to as a differential shape.
As illustrated in FIG. 14B, among the partial shapes constituting these font images, the first level pattern has a particularly high appearance frequency because of the common shape of these font images. The second level pattern a and the second level pattern b are partial shapes unique to the respective font images, and their appearance frequency is lower than that of the first level pattern.
Therefore, the image processing apparatus 2 according to the present modification hierarchically registers the common shape and the difference shape of these font images in the image dictionary as image patterns for the font image group classified for each character code.

FIG. 15 is a flowchart of the image pattern creation process (S310) in this modification. It should be noted that among the processes in this figure, the same reference numerals are assigned to the processes that are substantially the same as those shown in FIG.
As illustrated in FIG. 15, in S216, the dictionary determining unit 540 determines whether or not a region where the number of distributions is not 0 (black pixel region) is wider than the reference in the pixel distribution after the threshold processing. If it is, the process proceeds to S312. If it is narrower than the reference, the image pattern creation process (S310) ends without registering the image pattern for this font image group.
In step 312 (S312), the dictionary determining unit 540 extracts a portion larger than the first level threshold in the distribution probability Q ′ (x) as an image pattern having a common shape (that is, the first level pattern). Here, the first level threshold is a reference for separating the common shape that the font images belonging to the font image group have in common and other shapes. In this example, almost all of the first level threshold values belong to the font image group. Distribution probability when black pixels are present in the font image (0.95, etc.).

In step 314 (S314), the dictionary determination unit 540 sets the portion corresponding to the first level pattern to 0 from the distribution probability Q ′ (x) and removes the portion corresponding to the first level pattern from the distribution probability Q1 ( x).
Next, the dictionary determining unit 540 sets the portion of the distribution probability Q1 (x) equal to or higher than the second level threshold as 1 and sets the portion lower than the second level threshold as 0, and the sum coupling pattern Q1 from which the first level pattern has been removed. "(X) is generated. Here, the second level threshold is a standard for separating the difference shapes constituting the respective font images as the difference between the common shapes and having a high appearance frequency in the font image group." It is lower than the first level threshold, and is set according to the target code amount and image quality, for example.
Then, the dictionary determination unit 540 performs a product operation on the sum coupling pattern Q1 ″ and the pixel distribution P (i, x−vi) of each font image, and extracts Ri (x) that is a common part of these products. The extracted Ri (x) is determined to be an image pattern having a differential shape (that is, a second level pattern).

  In step 316 (S316), the dictionary determining unit 540 associates the first level pattern and the second level pattern generated from the same font image group (font images classified by the same character code) with each other, and creates an image dictionary. Register with. More specifically, the dictionary determining unit 540 registers the first level pattern and the second level pattern in the image dictionary in a hierarchical association with each index.

FIG. 16 is a diagram schematically illustrating the image pattern creation process (S310) in the present modification.
As shown in FIG. 16, the black pixel distribution probability Q ′ (x) of the font image classified by the character code indicates a different numerical value depending on the pixel position x. Of these, the portion with the highest distribution probability is considered to have a shape common to font images belonging to this font image group (that is, a common shape). The other area where the distribution probability exists is considered to have a shape corresponding to the difference between the respective font images and the common shape (that is, the difference shape).
Therefore, the dictionary determination unit 540 has a first level threshold for extracting a common shape (an area where the distribution probability is almost maximum) and a second level threshold for extracting a difference shape having a high appearance frequency. ing.
First, the dictionary determination unit 540 performs threshold processing on the distribution probability Q ′ (x) using the first level threshold, and extracts a portion corresponding to the first level pattern (common shape). Next, the portion corresponding to the extracted first level pattern is removed from the distribution probability Q ′ (x) and converted to “1” or “0” with the second level threshold as a reference, thereby A two-level sum coupling pattern Q1 ″ (x) is generated.
The dictionary determination unit 540 performs a product operation on the second level sum combination pattern Q1 ″ (x) and the pixel distribution P (i, x) of each font image i, thereby obtaining a second level which is a common part of these products. In this example, the common part between the second level sum combined pattern Q1 ″ (x) and “font image # 1” is the second level pattern a, and the second level sum combined pattern Q1 ″ (x ) And “font image # 2” are the second level pattern b.
Thereby, “font image # 1” can be replaced with the respective indexes as the sum of the first level pattern and the second level pattern a, and “font image # 2” can be replaced with the first level pattern and the second level pattern b. Can be replaced with an index as the sum of

FIG. 17 is a flowchart of the encoding process (S340) in this modification.
As shown in FIG. 17, in step 342 (S 342), the pattern determination unit 610 is registered in the image dictionary and a partial image (that is, a character image) cut out from the input image based on the corrected position information. The number of matching pixels is calculated by comparing with the first level pattern.
In step 344 (S344), the pattern determination unit 610 determines whether there is a first level pattern that matches the input partial image. Specifically, the pattern determination unit 610 determines whether or not the number of matching pixels calculated for each first level pattern is within an allowable range (for example, 90% or more with respect to all pixels of the partial image). If the value is within the allowable range, the process proceeds to S346. If the value is outside the allowable range, the process proceeds to S356.

In step 346 (S346), the pattern determination unit 610 reads from the image dictionary the index of the first level pattern having the largest number of matching pixels out of the first level patterns whose matching pixel number is within the allowable range, and is read out. The index is output to the index encoding unit 630, and the position information of the character image (corrected by the position correction unit 550) is output to the position information encoding unit 620.
In step 348 (S348), the pattern determination unit 610 extracts a difference shape between the shape of the first level pattern and the input partial image.
In step 350 (S350), the pattern determination unit 610 reads out the second level pattern associated with the first level pattern from the image dictionary, the read second level pattern, and the extracted difference shape. Are compared to calculate the number of matching pixels. The pattern determination unit 610 determines whether or not the number of matching pixels calculated for each second level pattern is within an allowable range (for example, 90% or more with respect to all pixels of the difference shape). If it is, the process proceeds to S352, and if it is outside the allowable range, the process proceeds to S354.

In step 352 (S352), the pattern determination unit 610 reads from the image dictionary the second level pattern index having the largest number of matching pixels from the second level pattern in which the number of matching pixels is within the allowable range, and is read out. The index is associated with the index of the first level pattern and output to the index encoding unit 630.
In this example, black pixels that do not match either the first level pattern or the second level pattern are removed as noise or the like, but these black pixels themselves may be encoded separately.

In step 354 (S354), the position information encoding unit 620 encodes the position information input from the pattern determination unit 610, and outputs code data of the position information to the selection unit 660.
The selection unit 660 associates the code data of the index input from the index encoding unit 630 and the code data of the position information input from the position information encoding unit 620 with each other and outputs them to the code output unit 670. . That is, the selection unit 660 outputs the index and the position information to the code output unit 670 so as to be associated with each other for each partial image.

In step 356 (S356), the pattern determination unit 610 outputs the partial image (that is, the character image in which the corresponding first level pattern does not exist in the image dictionary) to the image encoding unit 640.
The image encoding unit 640 encodes the partial image image data input from the pattern determination unit 610 and outputs the partial image code data to the selection unit 660.
The selection unit 660 outputs the code data of the partial image input from the image encoding unit 640 to the code output unit 670.

In step 358 (S358), the pattern determination unit 610 determines whether or not the encoding process has been completed for all the partial images, and when there is an unencoded partial image, the process returns to the process of S342, When the next partial image is encoded and all the partial images have been encoded, the process proceeds to S360.
In step 360 (S360), the dictionary encoding unit 650 encodes the image dictionary (the image pattern and the index associated with each other) input from the image dictionary creation unit 50, and outputs the code data of the image dictionary as code output. Output to the unit 670.

As described above, the image processing device 2 in the present modification compares a plurality of font images classified by the same character code with each other, extracts a common shape and a difference shape, and extracts the extracted common shape and The difference shapes are hierarchically associated and registered in the image dictionary. Thereby, the reproducibility of a character image can be improved, achieving a high compression rate.
In addition, the image processing apparatus 2 gradually improves the image quality according to the amount of code data decoded on the decoding side by hierarchizing the common shape and the difference shape and sequentially outputting the code data of the common shape. Progressive transmission can be realized.

Further, the image dictionary creation unit 52 may create an image dictionary according to the accuracy of the character recognition process. For example, the image dictionary creation unit 52 classifies the font images corresponding to the character images according to the accuracy of the character recognition processing for each of the clipped character images, and performs character recognition processing based on the classified font images. Create an image dictionary for each accuracy.
This is because the accuracy (accuracy) of character recognition by the character recognition unit 410 may differ for each character image included in the input image, and the font image group classified based on the character recognition result (character code) This is because character images having similar shapes may be mixed. That is, when character images that are not similar in shape are mixed in the character image group, an adverse effect such as distortion of the image pattern registered in the image dictionary may occur. An image dictionary is created for each accuracy of character recognition processing to minimize the influence of mistakes in the character recognition processing.

FIG. 18 is a diagram illustrating an image dictionary created for each accuracy of character recognition processing.
As illustrated in FIG. 18, the character classification unit 520 acquires the accuracy of the character recognition processing from the character recognition unit 410, and in accordance with the acquired accuracy, the font image (character image and character image) corresponding to the extracted character image. The best matching font image). The character classifying unit 520 in this example is a font image group with “accuracy is 90% or more”, a font image group with “accuracy is 70% or more and less than 90%”, and a font image group with “accuracy is less than 70%”. As described above, the font images are classified according to the accuracy range.
Further, the character classifying unit 520 further classifies each font image group with a character code.

The image dictionary determination unit 540 creates an image pattern based on the font image group classified by the accuracy and the character code. When determining an image pattern for a font image group with low accuracy, the dictionary determination unit 540 first determines an image pattern determined for a font image group with high accuracy and this font image group (font image with low accuracy). Each font image belonging to the group) is compared to determine whether or not they match, and if they match, the font images are excluded from the font image group.
As described above, the image processing apparatus 2 can minimize the influence on the image dictionary due to the mistake of the character recognition process by creating the image dictionary for each accuracy of the character recognition process.

(A) illustrates an encoding method on the assumption that a common font database exists, and (B) illustrates an encoding method on the assumption that an image dictionary is attached. (A) illustrates an image dictionary, and (B) illustrates an image pattern. It is a figure which illustrates the hardware constitutions of the image processing apparatus 2 with which the image dictionary creation method concerning this invention is applied centering on the control apparatus 20. FIG. It is a figure which illustrates the function structure of the encoding program 5 which is performed by the control apparatus 21 (FIG. 3), and implement | achieves the image dictionary creation method concerning this invention. It is a figure explaining the function of the 1st image dictionary creation part 50 in detail. It is a figure explaining the function of the encoding part 60 in detail. It is a flowchart which shows 1st operation | movement (S10) of the encoding program 5. It is a figure explaining the relationship between a character code and a font image. It is a figure explaining the function structure of the 2nd image dictionary preparation part. It is a flowchart which shows 2nd operation | movement (S20) of the encoding program 5. FIG. It is a flowchart explaining the image pattern creation process (S210) in 2nd Embodiment in detail. It is a figure which illustrates typically image pattern creation processing (S210) in a 2nd embodiment. It is a flowchart explaining the encoding process (S240) in 2nd Embodiment in detail. It is a figure which illustrates the common shape and difference shape of several character images which can be determined as the same character code. It is a flowchart of the image pattern creation process (S310) in this modification. It is a figure which illustrates typically image pattern creation processing (S310) in this modification. It is a flowchart of the encoding process (S340) in this modification. It is a figure which illustrates the image dictionary produced for every probability of character recognition processing.

Explanation of symbols

DESCRIPTION OF SYMBOLS 2 ... Image processing apparatus 5 ... Coding program 40 ... Image input part 410 ... Character recognition part 420 ... PDL decomposer 50,52 ... Image dictionary creation part 500 ... Storage part 510: Character image extracting unit 520: Font providing unit 525 ... Font database 530 ... Match determining unit 540 ... Dictionary determining unit 550 ... Position correcting unit 560 ... Index adding unit 570 ... Registration control unit 580 ... Character classification unit 60 ... Encoding unit 610 ... Pattern determination unit 620 ... Position information encoding unit 630 ... Index encoding unit 640 ... Image code Conversion unit 650... Dictionary encoding unit 660... Selection unit 670.

Claims (14)

A template providing means for providing a typical character image as a template image;
A match determination means for determining a match degree between the character image included in the input image and the template image provided by the template providing means;
A type determining unit that determines a template image provided by the template providing unit as a typical image pattern in the input image based on the degree of matching determined by the matching determining unit;
An image dictionary creating apparatus comprising: identification information providing means for assigning identification information for identifying each image pattern to the image pattern determined by the type determining means. The template providing means provides a font image as the template image,
The image dictionary creation device according to claim 1, wherein the matching determination unit determines the degree of matching by comparing a provided font image with a character image included in an input image. Font storage means for storing at least one font image in association with each character code;
An information acquisition means for acquiring a character code of a character image included in the input image;
The template providing unit reads a font image corresponding to the character code acquired by the information acquisition unit from the font storage unit, and provides the read font image as the template image.
The image dictionary creation device according to claim 2, wherein the matching determination unit determines the degree of matching by comparing a provided font image with a character image included in the input image. When the plurality of font images corresponding to the character code are provided by the template providing unit, the match determination unit compares each of the provided font images with the character image included in the input image, Determine the degree of match for each font image,
The image dictionary creation device according to claim 3, wherein the type determination unit selects, as an image pattern, a font image that most closely matches the character image based on the degree of matching determined for each font image. Information acquisition means for acquiring character area information indicating the area of each character image in the input image;
A character image extracting means for extracting a character image from the input image based on the character area information acquired by the information acquiring means;
3. The image dictionary according to claim 1, wherein the match determination unit determines the degree of match by comparing the character image extracted by the character image extraction unit with the template image provided by the template providing unit. Creation device. The match determination means compares the character image extracted by the character image extraction means with the template image provided by the template providing means at a plurality of relative positions, and determines the degree of match.
The image dictionary creation device according to claim 5, further comprising: a region correction unit that corrects the character region information of each character image based on the degree of matching determined at each relative position by the matching determination unit. Pattern storage means for storing image patterns sequentially determined by the type determination means;
The image pattern stored by the pattern storage means is compared with the newly input character image, and when any image pattern matches this character image, determination of the image pattern for this character image is prohibited. And a registration control means for
In accordance with an instruction from the registration control unit, the type determining unit is configured to provide a template provided for the character image when a newly input character image matches an image pattern stored in the pattern storage unit. When an image is not registered in the pattern storage unit and an input character image does not match any image pattern, a template image provided for the character image is registered in the pattern storage unit as an image pattern. Item 7. The image dictionary creation device according to any one of Items 1 to 6. Information acquisition means for acquiring character identification information for identifying a character represented by a character image included in the input image;
Character classification means for classifying the character images included in the input image into a plurality of character image groups based on the character identification information acquired by the information acquisition means;
A template providing means for providing a typical character image as a template image;
Match determination means for comparing each character image classified into the character image group by the character classification means and the template image provided for each character image group by the template providing means, respectively, for determining the degree of match. When,
A type determining unit that determines an image pattern based on the template image provided by the template providing unit and the degree of matching determined for the character image and the template image;
An image dictionary creating apparatus comprising: identification information providing means for assigning identification information for identifying each image pattern to the image pattern determined by the type determining means. The type determining means determines the appearance frequency in each character image group of the template image that most closely matches each character image based on the degree of matching determined by the match determining means, and based on the determined appearance frequency, The image dictionary creation device according to claim 8, wherein one or more image patterns are determined for a plurality of character images classified into the same character image group. The match determination means compares a plurality of character images classified into each character image group by the character classification means with a template image, determines a matching degree for each region included in the character image,
9. The image dictionary creation device according to claim 8, wherein the type determining unit determines a plurality of image patterns constituting a character image based on a matching degree determined for each region by the matching determining unit and a template image. . A template providing means for providing a typical character image as a template image;
A match determination means for determining a match degree between the character image included in the input image and the template image provided by the template providing means;
A type determining unit that determines a template image provided by the template providing unit as a typical image pattern in the input image based on the degree of matching determined by the matching determining unit;
Identification information giving means for giving identification information for identifying each image pattern to the image pattern determined by the type determining means;
Encoding means for encoding the character image data included in the input image by substituting the character region information indicating the region in the input image of the character image and the identification information of the image pattern corresponding to the character image And an encoding device. The match determination unit compares the character image included in the input image with the template image provided by the template providing unit at a plurality of relative positions, and determines the degree of match.
Information acquisition means for acquiring character area information indicating the area of each character image in the input image;
An area correction unit that corrects the character area information acquired by the information acquisition unit based on the degree of matching determined at each relative position by the match determination unit;
The encoding unit according to claim 11, wherein the encoding unit encodes the character region information of the character image corrected by the region correction unit and the identification information of the image pattern corresponding to the character image in association with each other. apparatus. Provide a typical character image as a template image,
Determine the degree of match between the character image included in the input image and the provided template image,
Based on the determined degree of matching, the provided template image is determined as a typical image pattern in this input image,
An image dictionary creation method for providing identification information for identifying each image pattern to a determined image pattern. In an image dictionary creation device including a computer,
Providing a typical character image as a template image;
Determining the degree of matching between the character image included in the input image and the provided template image;
Determining the provided template image as a typical image pattern in the input image based on the determined degree of matching;
A program for causing the computer of the image dictionary creating apparatus to execute identification information for identifying each image pattern with respect to the determined image pattern.

Images