JP2010191767A - Device and method for recognizing character - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To improve character recognizing precision by specifying the connection point of characters in either a horizontal direction and a vertical direction to correctly separate those characters. <P>SOLUTION: A character recognition method includes steps of: extracting the contours of a character string by using a differential filter from image data; calculating the sign of a differential value by the differential filter about each of elements configuring vertical contours and horizontal contours among the extracted contours; extracting a first region with a pixel under consideration as a center from among pixels configuring the vertical contours and the horizontal contours; calculating the number of pixels indicating the same signs and the number of pixels indicating opposite signs by comparing the signs of the pixels included in the first region with the sign of the pixel under consideration; determining the contrast of the first region based on the number of the same signs or the opposite signs; and define the pixels showing the signs contrary to the determination result in the first region as the division point of the character string. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a character recognition device that recognizes a character string such as a vehicle number, and more particularly to a character recognition device that can cut out characters with high accuracy even when the interval between adjacent characters is narrow. It is.

2. Description of the Related Art Conventionally, there has been known an apparatus for automatically recognizing description contents such as a vehicle number described on a license plate of a vehicle.
In order to recognize the description content of the license plate, a character recognition method using an image processing technique is used. In such a character recognition method, generally, after inputting a target character image, the character line is emphasized (binarization) to separate the character and the background, and an area that can be determined to be a character is extracted ( Character cutout process). In the character cut-out area, character shape features (for example, contour direction features) are extracted, and the characters are recognized by pattern recognition.

By the way, the license plate has a different shape, size, written character string, etc. for each country. For example, the character and the character are close to each other and appear to be connected to each other, or the license plate is attached to the vehicle. There are also some license plates where characters and peripheral structures such as bolts and the like are close to each other and characters appear to be connected. In such a case, the conventional character recognition device cannot perform a correct character cut-out process, and a correct recognition result cannot be obtained.
Therefore, Patent Document 1 discloses a technique for maintaining the recognition accuracy even when a character string to be described is a document in which Japanese and English are mixed, using a contour profile of binarized image data. .

JP 2001-22885 A

  However, in the technique described in Patent Document 1, when characters or peripheral structures are close to each other in the vertical direction of the character string to be recognized and the characters and the characters and the peripheral structures appear to be connected, There is a problem that it is difficult to specify the location, and the character recognition accuracy is lowered because the connected location of the characters cannot be correctly separated.

  The present invention has been made to solve the above problem, and provides a character recognition device that improves character recognition accuracy by identifying and correctly separating character connection points in both the horizontal and vertical directions. For the purpose.

  In order to solve the above problems, the present invention employs the following means.

  The present invention is a character recognition device that recognizes each character included in the character string by dividing the character string of image data obtained by imaging the character string, and using the differential filter from the image data A contour line extracting means for extracting the contour line, and a code calculating means for calculating a sign of a differential value by the differential filter for each pixel constituting a vertical contour line and a horizontal contour line among the extracted contour lines; One pixel is selected as a target pixel from the pixels for which the code has been calculated, a first area including the target pixel is sequentially extracted from the image data, and the code calculation is performed for pixels included in the first area. Based on the code calculated by the means, it is determined whether or not there is a pixel having a different sign from the sign of the pixel of interest, and there is a pixel having a different sign in the first region extracted from the image data The code change calculating means for calculating the number of the determined first areas, the first area determined to have a pixel having a different sign, the area where the target pixel indicates a + sign, and the target pixel are − There is provided a character recognition device, comprising: a dividing point determination unit that classifies a region indicating a code, and uses a smaller number of regions as a result of the separation to determine a dividing point of the character string.

  According to the present invention, a contour line of the character string is extracted from image data using a differential filter, and a differential value obtained by the differential filter is determined for each pixel constituting a vertical contour line and a horizontal contour line among the contour lines. Since the code is calculated, the luminance gradient can be grasped for both the vertical contour line and the horizontal contour line. Then, one pixel is selected as the target pixel from the pixels for which the codes are calculated, and the first region including the target pixel is sequentially extracted from the image data. That is, in order from the top of the image data, one pixel among the pixels whose codes are calculated is selected as a target pixel, and a first region based on the target pixel is extracted. Here, the first area is determined based on the size, thickness, etc. of the character to be recognized. If the character to be recognized is, for example, a character described on a license plate of a vehicle, the first area is 4. It is preferable to define the area of about ˜5 pixels. Then, with respect to the pixels included in the first region, it is determined whether there is a pixel that has a code different from the code of the pixel of interest based on the previously calculated code, and the first extracted from the image data. The number of first regions determined to have pixels having different signs in one region is calculated. This process is performed on all the first areas sequentially extracted from the entire image data. By doing so, it is possible to grasp the change in the sign with respect to the target pixel in the first region, and to determine the brightness of the first region based on the calculation result. In the determination of lightness and darkness, the first region determined to have a pixel indicating a sign is classified into a region where the target pixel is positive and a region where the target pixel is negative. An area is set as a division point of the character string. That is, the number of areas where the pixel of interest has a + sign while the area of the area has a − sign pixel and the area of interest where the pixel of interest is − sign and the area having a + sign pixel in the area In comparison, a larger area is an area that captures a character line, and a smaller area is a segment that divides a character string as an area that captures a background.

  In this way, since the brightness of the character recognition target pixel is determined based on both the vertical and horizontal gradients of the contour line, the connected portion of the characters and peripheral structures such as bolts for fixing the character and the license plate, Can be correctly identified, and the connected portion can be correctly divided. This improves the character recognition accuracy.

  The character recognition device includes a region selection unit that preliminarily estimates a position of the division point and determines a second region including the estimated division point position from the image data, and the division point determination unit includes: The division point of the character string may be determined in the second area.

  Since the position of the dividing point is estimated in advance and the second area including the position of the dividing point estimated from the image data is determined, it is only necessary to execute the character recognition process within the selected area, which is required for the character recognition process. Time can be shortened and erroneous detection of division points can be suppressed.

  The character recognition device includes voting means for generating a vote distribution consisting of peripheral distributions that have been projected and voted by voting on a voting space in which the division points are projected in a horizontal axis direction, and the division point determination means includes the From the vote distribution, when the vote value in the horizontal direction is equal to or greater than a predetermined threshold, the vote value in the vertical direction with respect to the vote value in the horizontal direction may be set as a true division point.

  By voting on the voting space in which the division points are projected in the horizontal axis direction, a vote distribution consisting of the peripheral distribution voted on by projection is generated, and when the vote value in the horizontal direction is greater than or equal to a predetermined threshold, Since the vertical voting value with respect to the directional voting value is set as the true dividing point, even in the case of image data including a character string, when there are many connected parts such as when characters and the peripheral structure are in close contact with each other Occurrence of division leakage can be prevented and character recognition accuracy is improved.

  The character recognition device includes a storage unit that stores character information related to the size of the character in advance, and the division point determination unit compares the character information with the division point and displays the character indicated by the character information. The division point closest to the size of may be the true division point.

  Character information related to the size of the character is stored in the storage means in advance, and the character information and the division point are compared, and the division point closest to the character size indicated by the character information is set as the true division point. Since the division point is determined based on the information related to the brightness gradient of the outline and the information related to the character size, the character recognition accuracy is improved.

  The present invention is a character recognition method for dividing the character string of image data obtained by imaging a character string and individually recognizing characters included in the character string, and using the differential filter from the image data, the character string A step of extracting a contour line, a step of calculating a sign of a differential value by the differential filter for each pixel constituting a vertical contour line and a horizontal contour line among the extracted contour lines, and the sign is calculated. One pixel is selected as the target pixel from the selected pixels, the first region including the target pixel is sequentially extracted from the image data, and the pixels included in the first region are calculated by the code calculation unit Based on the code, it is determined whether or not there is a pixel having a different sign from the code of the target pixel, and it is determined that there is a pixel having a different sign in the first region extracted from the image data. The step of calculating the number of first regions, and the first region determined to have a pixel having a different sign are classified into a region in which the target pixel indicates a + sign and a region in which the target pixel indicates a-sign. There is provided a character recognition method comprising the steps of: classifying, and, as a result of the classification, a region having a smaller number as a division point of the character string.

  As described above, according to the present invention, since the connection point of characters can be correctly specified in both the horizontal direction and the vertical direction, the connection point can be correctly separated and the character recognition accuracy can be improved.

It is a block diagram which shows schematic structure of the character recognition apparatus concerning the 1st Embodiment of this invention. It is a flowchart which shows the process of the character recognition process in the character recognition apparatus concerning the 1st Embodiment of this invention. It is explanatory drawing which showed the process of the character recognition process, (a) is the image | photographed character string, (b) is the character string by which the outline enhancement process was made, (c) is the vertical edge image of a character string, (D) shows the horizontal edge image of a character string, respectively. FIG. 4 is an enlarged view of a portion A in FIG. 3 (c), and is an explanatory diagram when a code is calculated in the character recognition processing of the present invention. It is a flowchart which shows the process of the character recognition process concerning the 2nd Embodiment of this invention. It is a flowchart which shows the process of the character recognition process concerning the 3rd Embodiment of this invention.

Hereinafter, embodiments of a character recognition device according to the present invention will be described with reference to the drawings.
[First Embodiment]
FIG. 1 is a block diagram showing a schematic configuration of a character recognition apparatus according to the first embodiment of the present invention.

  The character recognition device 10 according to the present invention includes a camera 11 that captures a license plate of a vehicle and obtains image data, an image input unit 12 that receives input of image data captured by the camera 11, and an image input unit 12 A recognition processing unit 13 that executes character recognition processing based on image data and a recognition result output unit 14 that outputs a recognition result in the recognition processing unit 13 are provided.

  The recognition processing unit 13 performs calculations for executing the character recognition process, and stores a CPU (central processing unit) 21 that executes various calculation processes related to the character recognition process, a program related to the character recognition process, and the like. A ROM (Read Only Memory) 22 which is a read-only memory, a RAM (Random Access Memory) 23 which is a readable / writable memory functioning as a work area of the CPU 21, and a character recognition process based on a program related to the character recognition process are executed. In this case, a storage device 24 is provided in which various data necessary for this are stored.

  In addition, the recognition processing unit 13 includes a contour line extraction unit 25 as a contour line extraction unit, a code calculation unit 26 as a code calculation unit, a code change calculation unit 27 as a code change calculation unit, and a division point determination for determining a division point. A division point determination unit 28 is provided as a means.

  The contour line extraction unit 25 calculates a spatial first derivative of the image data by using a first-order differential filter such as Sobel differentiation with respect to the image data input by the image input unit 12, and calculates the character in the image data. An outline is extracted (outline enhancement processing). In addition, the contour line is thinned by suppressing non-maximum contour pixels in the image data that has undergone the contour line emphasis process (contour line thinning process). The code calculation unit 26 receives image data that has undergone the contour extraction process and the contour thinning process, and the pixels constituting the vertical contour line and the horizontal contour line with respect to the processed image data, that is, the horizontal direction or The sign (+ or-) of the differential value of the contour is obtained for a pixel having a gradient in the vertical direction. Based on the code calculated by the code calculation unit 26, the code change calculation unit 27 extracts a local region as a first region corresponding to the thickness of the character from the image data, and the differential value of the pixel in the local region Is calculated. The dividing point determination unit 28 determines the lightness / darkness of the character with respect to the background color of the license plate based on the change of the sign calculated by the sign change calculation unit 27, and changes the sign opposite to the lightness / darkness of the character. The center of the local area is obtained, and this center is calculated as a dividing point. In the contour line extraction unit 25, the code calculation unit 26, the code change calculation unit 27, and the division point determination unit 28, the CPU 21 expands the processing program stored in the predetermined ROM 22 in the RAM 23 and executes the expanded program. The processing unit realized by the above-described processing will be described later.

  Hereinafter, the character recognition process in this embodiment is demonstrated with reference to FIG. 2, FIG. FIG. 2 is a flowchart showing the process of character recognition processing in the present embodiment.

  As shown in FIG. 2, the license plate image captured by the camera 11 is input to the image input unit 12 as the image data shown in FIG. 3A, and the recognition processing unit 13 performs predetermined processing in steps S101 and S102. Perform the process. That is, in step S101, the contour line extraction unit 25 calculates the spatial first derivative of the image data with respect to the input image data using a differential filter such as a Sobel filter, and as shown in FIG. A contour line of characters in image data is extracted (contour emphasis processing). In step S102, the contour line extraction unit 25 thins the contour line by suppressing the non-maximum contour pixel with respect to the image data subjected to the contour line emphasis process in step S101 (contour line thinning process). ).

  That is, as shown in FIG. 3C, a pixel having a gradient in the horizontal direction is extracted from the image data that has been subjected to the contour emphasis processing in step S101, and the contour direction is calculated (vertical edge image). Generation), a point (maximum point) at which the contour luminance value becomes a peak value in a direction perpendicular to the contour is searched, and this is extracted as a point on the true contour. Similarly, as shown in FIG. 3D, for the image data subjected to the contour emphasis processing in step S101, pixels having a gradient in the vertical direction are extracted to calculate the direction of the contour (horizontal Edge image generation), a point (maximum point) at which the contour luminance value reaches a peak value in a direction perpendicular to the contour line is searched, and this is extracted as a point on the true contour line.

  Subsequently, in step S103, the code calculation unit 26 of the recognition processing unit 13 applies the contour line to pixels having a gradient in the horizontal direction or the vertical direction with respect to the image data that has undergone the contour line emphasis processing and the contour line thinning processing. Find the sign of the derivative of. More specifically, the code calculation unit 26 pays attention to the gradient direction of the contour line for each pixel of the extracted contour line in the image data that has undergone the contour line emphasis process and the contour line thinning process. The sign of the differential value of the vertical contour of the differential filter is calculated for a pixel having a gradient in the vertical contour, that is, the vertical contour. Similarly, the sign of the differential value of the horizontal contour of the differential filter is calculated for a pixel having a gradient in the vertical direction, that is, the horizontal contour. Here, FIG. 4 is an enlarged view of a portion A in FIG. 3C, and shows an example in the case of calculating a code.

  In step S104, the sign change calculation unit 27 extracts a local area corresponding to the thickness of the character centered on the target pixel based on the sign of the differential value calculated in each pixel in step S103, and is included in the local area. It is calculated whether the sign of the pixel to be changed is compared with the sign of the target pixel. Specifically, first, the image data for which the sign of the differential value has been calculated in step S103 is sequentially scanned from the upper left pixel one line at a time in the right direction to extract the pixel for which the differential value has been calculated. . Then, among the pixels included in one scanned row, a pixel whose differential value is calculated is used as a target pixel, and a predetermined local region including pixels corresponding to the character thickness is extracted from the target pixel in the right direction. Subsequently, the local region is searched for a pixel having a sign opposite to the target pixel. If there is a pixel having a sign opposite to the target pixel, the sign changes with respect to the target pixel. As a region where a pixel is located, information on the coordinates of the pixel and the direction of sign change (a sign for the target pixel + code or a + sign for the target pixel-code) is stored. This process is performed over one scanned line, and when the process for this one line is completed, the process for the next line is started. The above processing is sequentially performed for each line, so that the processing is repeated for the entire image data.

  Then, from the calculated code change, the number of regions where the pixel whose code has changed with respect to the target pixel is calculated for each piece of information of the code change direction. That is, the number of areas where there is a pixel that changes to − sign when the target pixel is + sign, and the number of areas where there is a pixel that changes to + sign when the target pixel is − sign, respectively. calculate.

  Here, the character thickness equivalent is often an area corresponding to about 4 to 5 pixels, and the range of the local area is the size, thickness, etc. of the character written on the license plate to be recognized. Is stored in advance in the storage device 24 as data determined in consideration of the above. In addition, when calculating horizontal division points, a local region is determined in the horizontal direction to calculate a change in sign, and when dividing points in the vertical direction are calculated, a local area is determined in the vertical direction and the code is changed. Calculate the change.

  In step S105, the division point determination unit 28 determines whether the local area is bright or dark with respect to the background, that is, whether the local area is an area covering the character line portion or an area covering the background portion, from the calculation result in step S103. To do. Specifically, by comparing the number of regions where there is a pixel that has changed to −sign with respect to the target pixel + sign and the number of regions with pixels that have changed to + with respect to the target pixel− It is determined that the local region of the larger local region captures the character line portion. For example, if there are more regions with pixels that have changed to + sign with respect to the pixel of interest-sign than there are regions with pixels that have changed to-sign with respect to the pixel of interest + code, A region having a pixel that has changed to a + sign with respect to the pixel-code is determined as a character line portion, and a region having a pixel that has changed to a-sign with respect to the target pixel + code is determined as a background portion.

  In the next step S106, the dividing point determination unit 28 determines the dividing points in the local region. Specifically, the dividing point determination unit 28 determines the determination result of the local region related to the character line portion determined in step S105 and the local region related to the background portion, and the target pixel in the local region calculated in step S104. All the local regions determined as background parts are extracted based on the coordinates of the pixels whose sign is changed and information on the direction of the sign change. Then, the central pixel of the extracted local area is determined as a division point.

  When the division points are determined, the characters written on the license plate are separated, and character recognition processing is performed on individual character images cut out by binarization, labeling, etc. for the separated one character area Is made. When the predetermined character recognition process is completed for all the character areas, the recognition result is output to the recognition result output unit 14, and the character recognition process is completed.

  As described above, according to the present embodiment, since the brightness of the character recognition target pixel is determined based on both the vertical and horizontal luminance gradients of the contour line, the connection portion between characters and the character and license plate are fixed. It is possible to correctly identify the connection location with surrounding structures such as bolts to be connected, and to correctly divide the connection portion. This improves the character recognition accuracy.

  In addition, the location where a character and the connection of a character and a surrounding structure may arise is decided to some extent by the license plate of recognition object. For this reason, the position of the dividing point is estimated in advance by using information on the size of the character by an area selection unit (not shown), and the area including the estimated position is determined as an area for searching for the dividing point. In this way, the above-described process only needs to be executed within a predetermined area, and the time required for the character recognition process can be shortened and erroneous detection of division points can be suppressed. . The process for determining the search target area from the input image data by the area selection means can be performed at an arbitrary timing. For example, it can be performed prior to the brightness determination of the target pixel. It can also be done in advance. The area for searching for the dividing point may be calculated and selected every time the character recognition process is performed, but may be determined in advance in the storage means or the like and read from the storage means by the area selection means. The area selecting means also functions as a processing unit realized by the CPU 21 developing a processing program stored in a predetermined ROM 22 in the RAM 23 and executing the expanded program.

[Second Embodiment]
Next, a second embodiment of the present invention will be described with reference to FIG.

  The character recognition process of the present embodiment is different from that of the first embodiment described above. After the division point calculation process in the first embodiment is executed, the division point obtained here is set as a division candidate point, and the division candidate is obtained. Projection voting on the horizontal axis (x-axis) of a point is performed, and for coordinates where a large number of division points are voted, processing is performed to calculate all y coordinates of the x-coordinates as division points. Hereinafter, regarding the character recognition processing of the present embodiment, description of points that are common to the first embodiment will be omitted, and different points will be described.

  FIG. 5 is a flowchart showing a process of character recognition processing in the present embodiment. As shown in FIG. 5, the license plate image taken by the camera 11 in steps S201 to S206 is input to the image input unit 12 as image data shown in FIG. In the same manner as the character recognition process in the first embodiment described above, the number of regions in which pixels are changed to the minus sign with respect to the local attention pixel + code, and the change to the plus sign with respect to the noticeable pixel-code. A pixel having a smaller number of regions is extracted as a division candidate point.

  In step S207, voting means (not shown) generates a vote distribution consisting of the peripheral distributions that have been projectively voted by voting on the voting space obtained by projecting the division candidate points calculated in step S206 in the horizontal direction (x-axis direction). To do. In step S208, the x coordinate at which the vote is equal to or greater than a predetermined threshold is calculated from the generated vote distribution, and the process proceeds to the next step. In addition, as the predetermined threshold, when the average value of the maximum vote value or the total vote value is equal to or more than a predetermined fixed value, a value obtained by multiplying the maximum vote value by a predetermined arbitrary ratio, A value obtained by multiplying a reference height of a character predetermined by information by a predetermined arbitrary ratio can be used. In step S209, for all x coordinates calculated in the previous step S208, all y coordinates of the x coordinates are calculated, this y coordinate is determined as a division point, and this routine is terminated.

  When the division points are determined, the characters written on the license plate are separated, and character recognition processing is performed on individual character images cut out by binarization, labeling, etc. for the separated one character area Is made. When the predetermined character recognition process is completed for all the character areas, the recognition result is output to the recognition result output unit 14, and the character recognition process is completed.

  As described above, according to the present embodiment, since the brightness of the character recognition target pixel is determined based on both the vertical and horizontal luminance gradients of the contour line, the connection portion between characters and the character and license plate are fixed. It is possible to correctly identify the connection location with surrounding structures such as bolts to be connected, and to correctly divide the connection portion. In particular, by performing projective voting, it is possible to prevent the occurrence of division leakage even when there are many connected parts, such as when characters and their surrounding structures are in close contact with each other, and character recognition accuracy is improved. .

[Third Embodiment]
Subsequently, a third embodiment of the present invention will be described with reference to FIG.

  The difference between the character recognition processing of the present embodiment and the first embodiment and the second embodiment described above is that the recognition target is obtained after the division point calculation processing in the first embodiment and the second embodiment is executed. The point of division is further narrowed down using a priori information on the size of the character, character recognition processing is performed in the narrowed down region, and the region with the highest matching degree is determined as the character dividing point. Description of points common to the first and second embodiments will be omitted, and different points will be described.

  FIG. 6 is a flowchart showing the process of character recognition processing in this embodiment. As shown in FIG. 6, the division points calculated according to the first embodiment or the second embodiment described above and the character information that is information related to the character size are input to the recognition processing unit 13, and the character is The recognition process starts. The character information is information predetermined based on the license plate to be recognized, and the character information is stored in the storage device 24, for example.

  In step S301, the division candidate points are narrowed down based on the character information input from the storage device 24. That is, the character information is compared with all the division points calculated in the previous step, and the division candidate point closest to the character size indicated by the character information is extracted. In the next step S302, character recognition processing is performed on each character image by binarization, labeling, and the like at the extracted division candidate points. In the next step S303, the character that has been subjected to the character recognition process in the previous step S302 is compared with character information, and a combination of division candidate points having the highest matching degree is calculated.

  In step S304, it is determined whether or not the processing in steps S302 to S303 described above has been executed for all of the candidate division points narrowed down in step S301. If it is determined that the processing has not been executed, step S304 is executed. Returning to S302, the processes in steps S302 and S303 are repeated. If it is determined in step S304 that the processes in steps S302 to S303 have been executed for all of the division candidate points narrowed down in step S301, combinations of all the division candidate points calculated are adopted. The character dividing point is determined and this routine is terminated. Note that after the end of this routine, in order to make the character recognition result more reliable, it is also possible to perform character recognition processing again on individual character images cut out by processing such as binarization and labeling. . In this case, it is preferable to use parameters different from those of the character recognition process performed previously.

  As described above, according to the present embodiment, the character division point is determined based on the information related to the character gradient and the character matching result in addition to the information related to the luminance gradient of the character outline. Therefore, the character dividing point is determined more accurately, and the character recognition accuracy is improved.

DESCRIPTION OF SYMBOLS 10 Character recognition apparatus 11 Camera 12 Image input part 13 Recognition processing part 14 Recognition result output part 25 Outline extraction part 26 Code calculation part 27 Code change calculation part 28 Division | segmentation point determination part

Claims (5)

A character recognition device that divides the character string of image data obtained by imaging a character string and individually recognizes characters included in the character string,
Contour extraction means for extracting a contour line of the character string from the image data using a differential filter;
Code calculating means for calculating a sign of a differential value by the differential filter for each pixel constituting a vertical contour line and a horizontal contour line among the extracted contour lines;
One pixel is selected as a target pixel from the pixels for which the code has been calculated, a first area including the target pixel is sequentially extracted from the image data, and the code calculation is performed for pixels included in the first area. Based on the code calculated by the means, it is determined whether or not there is a pixel having a different sign from the sign of the pixel of interest, and there is a pixel having a different sign in the first region extracted from the image data Sign change calculating means for calculating the number of first regions determined as:
As a result of classifying the first area determined to have a pixel having an opposite sign into an area in which the pixel of interest indicates + sign and an area in which the pixel of interest indicates-sign, the number of areas is smaller. A dividing point determining means that defines a character string dividing point as
A character recognition device comprising: A region selection unit that preliminarily estimates the position of the dividing point and determines a second region including the estimated position of the dividing point from the image data,
The character recognition device according to claim 1, wherein the division point determination unit determines a division point of the character string in the second region. Voting means for generating a vote distribution composed of peripheral distributions that have been projected and voted by voting in a voting space in which the division points are projected in the horizontal axis direction;
The division point determining means, when the vote value in the horizontal direction is equal to or greater than a predetermined threshold from the vote distribution, sets the vertical vote value with respect to the horizontal vote value as a true division point.
The character recognition device according to claim 1, wherein the character recognition device is a character recognition device. Storage means for storing character information related to the size of the character in advance;
The dividing point determination means compares the character information with the dividing point and sets the dividing point closest to the character size indicated by the character information as a true dividing point.
The character recognition device according to claim 1, wherein the character recognition device is a character recognition device. A character recognition method for dividing the character string of image data obtained by imaging a character string and individually recognizing characters included in the character string,
Extracting a contour line of the character string from the image data using a differential filter;
Calculating a sign of a differential value by the differential filter for each pixel constituting a vertical contour line and a horizontal contour line among the extracted contour lines;
One pixel is selected as a target pixel from the pixels for which the code has been calculated, a first area including the target pixel is sequentially extracted from the image data, and the code calculation is performed for pixels included in the first area. Based on the code calculated by the means, it is determined whether or not there is a pixel having a different sign from the sign of the pixel of interest, and there is a pixel having a different sign in the first region extracted from the image data Calculating the number of first regions determined as:
As a result of classifying the first area determined to have a pixel having an opposite sign into an area in which the pixel of interest indicates + sign and an area in which the pixel of interest indicates-sign, the number of areas is smaller. And a step of defining the character string dividing point;
A character recognition method comprising:

Images