JPH08263588A - Character recognition device - Google Patents

Abstract

PURPOSE: To provide the character recognition device which can set a proper area to be recognized at a high speed. CONSTITUTION: When image data are stored in an image data storage part 1, a character line detection part 2 detects line spacing, and finds and stores line position coordinates in a character coordinate storage part 4. Also, a character cutting-out part 3 cuts out a character one by one for every line, and finds and stores character position coordinates in the storage part 4. When an operator specifies the area to be recognized through an area specification part 5, a recognized area determination part 6 calculates the coordinates closest to the specified coordinates among coordinate data stored in the character coordinate storage part 4 and stores them in a recognized area storage part 7. An area display means 8 generates an area frame from the coordinates of the area to be recognized stored in the recognized area storage part 7 and overlays it on the image data at a display part 9 in real time. Further, a recognition part 10 recognizes the characters in the final set recognized area.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character recognition device, and more particularly, to a character recognition device for recognizing a character image within a bit map image within an image area set based on an instruction from an operator. Is.

[0002]

2. Description of the Related Art A conventional character recognition device for recognizing a character image in an image area designated by an operator from a bitmap image is described in, for example, JP-A-4-5779 and JP-A-5-342620. Technology. In the technique described in Japanese Patent Laid-Open No. 457979/1992, when the operator designates the recognition target area by the input unit, if the character in the area is missing, the boundary feature extraction unit extracts it and changes the area. The recognition target area was changed by the department. When changing the recognition target area, set some boundary candidate lines parallel to the side where the character is missing, count the black pixels on the boundary candidate line for each boundary candidate line, and check the boundary where the character is not missing. Select the candidate line that is closest to the original side. In such a method, the image data is accessed after the operator specifies the recognition target area, which takes time. Further, there is also a problem that the recognition target area cannot be set when there is no candidate whose character is not missing in the boundary candidate line.

Further, in the technique disclosed in Japanese Patent Laid-Open No. 5-342320, the operator has only one of the recognition target areas.
The point is designated, and the operator changes the frame displayed by the designation to set a desired recognition target area. In this technique, the operator needs to change the size of the frame while checking the state of the frame display and being careful not to cover the character with the frame, which imposes an excessive burden on the operator.

[0004]

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and when an operator designates a recognition target area, it is possible to set an appropriate recognition target area at high speed. An object of the present invention is to provide a character recognition device that can be used.

[0005]

SUMMARY OF THE INVENTION The present invention is a character recognition device for recognizing a character image in an image area set based on an instruction from an operator, and image storage means for storing image data that can be a target of character recognition. A coordinate data detecting means for detecting coordinate data relating to coordinates of a line and one character or each character component in the image data stored in the image storing means; and the coordinate detected by the coordinate data detecting means. Coordinate data storage means for storing data, area designation means for accepting designation of an area by an operator, and the coordinate data stored in the coordinate data storage means that is closest to the coordinate designated by the area designation means. It is characterized by having a recognition area determining means for calculating coordinates.

According to a second aspect of the present invention, in the character recognition apparatus according to the first aspect, the area frame of the recognition area specified by the coordinates determined by the recognition area determining means is in contact with the character image. An area frame contact detection means for detecting whether or not there is an area frame contact detection means; and an area frame correction means for expanding or reducing an area frame contact portion of the character image detected by the area frame contact detection means so as not to overlap a character. It is characterized by.

[0007]

According to the present invention, from the image data stored in the image storage means, the coordinate data detection means detects beforehand the coordinate data relating to the coordinates of each line and one character or each character component, and the coordinate data storage means. Remember. Upon receiving the designation of the area from the operator, the recognition area determination means calculates the coordinates closest to the coordinates specified by the operator in the coordinate data stored in the coordinate data storage means, and determines the recognition area. Thus, even if the operator specifies the coordinates on the character when specifying the recognition area, the coordinates that do not overlap the character are calculated and the recognition area is set. Further, when calculating coordinates such that a character to be recognized is not missing, only coordinate data need be processed without referring to image data, and high-speed processing is possible.

Further, whether or not the area frame of the recognition area specified by the coordinates determined by the recognition area determination means is in contact with the character image in the image data is detected by the area frame contact detection means, and if it is recognized, When the area frame of the area is in contact with the character image, the area frame correcting means expands or reduces the area frame of the contacting area so as not to overlap the character. As a result, the recognition area is not set by omitting a part of the character to be recognized. Even if the area frame that does not overlap the character cannot be set near the area frame of the recognition area specified by the operator, the area frame can be set by expanding the area frame only for the part that overlaps the character. Becomes The area frame correction means may be configured to automatically correct the area frame, or may be expanded or reduced by an operator.

[0009]

1 is a block diagram showing the first embodiment of the character recognition apparatus of the present invention. In the figure, 1 is an image data storage unit, 2 is a character line detection unit, 3 is a character cutout unit, 4 is a character coordinate storage unit, 5 is an area designation unit, 6 is a recognition area determination unit, 7 is a recognition area storage unit, Reference numeral 8 is an area frame display unit, 9 is a display unit, and 10 is a recognition unit. The image data storage unit 1 stores image data including a character image to be recognized. The character line detection unit 2 refers to the image data stored in the image data storage unit 1 to detect the area where the character line exists,
The coordinates are acquired as the row position coordinates. The character cutout unit 3 refers to the image data stored in the image data storage unit 1, cuts out each character, and acquires the coordinates as the character position coordinates. At this time, the line position coordinates which can be detected by the character line detection unit 2 and stored and stored in the character coordinate storage unit 4 may be used. The character coordinate storage unit 4 stores the line position coordinates and the character position coordinates acquired by the character line detection unit 2 and the character cutout unit 3.

The area designating section 5 receives at least an instruction regarding the recognition area from the operator, and transmits the designated coordinates to the recognition area determining section 6. The recognition area determination unit 6 determines, based on the line position coordinates and the character position coordinates stored in the character coordinate storage unit 4, the line position coordinates or the character position closest to the coordinates of the recognition area specified by the area specifying unit 5. The coordinates are calculated and the recognition area is determined. The recognition area storage unit 7 stores the recognition area determined by the recognition area determination unit 6.

The area frame display unit 8 generates information for displaying the area frame based on the information of the recognition area stored in the recognition area storage unit 7 and causes the display unit 9 to display the information. The display unit 9 displays at least the image data stored in the image data storage unit 1, and is used when the operator designates the recognition region in the region designation unit 5. The operator can confirm the determined recognition area by displaying the area frame generated by the area frame display unit 8 so as to overlap with the image data.

The recognizing unit 10 operates the image data storing unit 1 according to the information of the recognizing region stored in the recognizing region storing unit 7.
The image of the recognition area is acquired from the image data stored in the, and the character image included therein is recognized.

FIG. 2 is a block diagram showing an example of the character line detection unit 2 and the character cutout unit 3. 11 in the figure
Is a horizontal black pixel number counting unit, 12 is a line spacing determining unit, 1
Reference numeral 3 is a vertical black pixel number counting unit, and 14 is an intercharacter determining unit. For example, if the characters are described in horizontal writing, it can be detected as an area between the lines in which there are almost no black pixels in the horizontal direction. Here, a row is detected by scanning the image data in the horizontal direction to obtain a projection. In the horizontal black pixel number counting unit 11 of the character line detecting unit 2,
The image data stored in the image data storage unit 1 is scanned in the horizontal direction, and the number of black pixels is counted for each horizontal line to create a projection. The line space determination unit 12 uses, for example, the projection created by the horizontal black pixel number counting unit 11 to calculate, for example,
Each character line is detected by comparing with a predetermined threshold value or by checking a change in the number of black pixels.

Similarly, there are almost no black pixels between the characters in each line. The vertical black pixel number counting unit 13 of the character cutting unit 3 vertically scans the image data in each line detected by the character line detecting unit 2 to count black pixels and create a projection. Then, based on the created projection in the vertical direction, the character spacing determination unit 14 determines the spacing between characters and detects each character. The horizontal black pixel number counting unit 11 and the vertical direction black pixel number counting unit 13 are configured to count the number of black pixels of the image data, but may be configured to count the number of white pixels. , The same detection can be performed.

FIG. 3 is a flow chart showing an example of the operation in the first embodiment of the character recognition device of the present invention.
In S21, the image data to be recognized is stored in the image data storage unit 1, and the stored image data is visualized by the operator on the display unit 9.

When the image data is stored in the image storage means 9, first in S22, the character line detection unit 2 detects the line position coordinates where each character line exists, and the character coordinate storage unit 4 is detected.
Stored in. For example, in the configuration shown in FIG. 2, in order to detect each character line, the line position coordinates can be obtained by scanning the image data in the horizontal direction and taking a projection to find the coordinates of the boundary between the black pixel and the white pixel. it can. Also, the character cutting means 3
Refers to the line position coordinates stored in the character coordinate storage means 4, cuts out a character area for each line, and stores the coordinates of the cut out character area in the character coordinate storage unit 4. For example, in the configuration shown in FIG. 2, the character position coordinates can be obtained by taking a vertical projection on the image data for one line and finding the coordinates of the boundary between the black pixel and the white pixel. In this way, the coordinates indicating the position of each character can be obtained. The coordinate detection processing by the character line detection unit 2 and the character cutout unit 3 may be performed in parallel while synchronizing with each character line.

At step S23, the operator operates the display unit 9
While looking at the image data displayed in, the coordinates of the starting point of the area to be recognized are set using the pointing device of the area designating unit 5 or the like. In S24,
When the coordinates of the starting point specified by the area specifying unit 5 are input to the recognition area determining unit 6, the recognition area determining unit 6 selects the specified coordinates from the coordinate data stored in the character coordinate storage unit 5. The coordinate data closest to is extracted by distance calculation, and the extracted coordinates are output to the recognition area storage unit 7 as the coordinates of the start point. Since the recognition area has not been determined at this point, the recognition target area is not displayed in S25. In S26, it is determined whether or not it is the last coordinate data. Since the coordinates of the end point have not been input yet, the process returns to S23.

In order to specify the coordinates of the other end point of the recognition area, the operator operates the area designating unit 5 while confirming the recognition target area on the display unit 9 in S23.
When the operator is operating the area designating unit 5, the display unit 9 outputs the coordinates currently designated by the area designating unit 5 at regular time intervals. In step S24, the recognition area determination unit 6 determines the coordinates designated by the area designation unit 5 at that time as the coordinate data closest to the designated coordinates among the coordinate data stored in the character coordinate storage unit 4. It is extracted by distance calculation and the extracted coordinates are output to the recognition area storage unit 7 as the coordinates of the end point. The coordinates of the start point and the end point output to the recognition area storage unit 7 are acquired by the area frame display unit 8, an area frame indicating the recognition area is generated, and passed to the display unit 9. In S25, the display unit 9 displays the area frame over the image data.

In this way, the recognition target area to be instructed by the operator is an area frame as a rectangular area having the start point and the coordinate pointed at that time as the end point until the final coordinates are input. It is displayed on the display unit 9 by the display unit 8. At this time, the area frame has been changed by the recognition area determination unit 6 to the coordinate data closest to the instructed coordinate among the coordinate data stored in the character coordinate storage unit 4, so that the display unit 9 displays the frame. The displayed area frame is displayed so that its start point and end point do not overlap with nearby characters.

When the coordinates are finally determined, the process proceeds to S27, and the recognition section 10 refers to the recognition area coordinates stored in the recognition area storage section 7 and stores them in the image data storage section 1. The character recognition process is executed on the image in the recognition area of the image data.

The above operation will be described using a specific example. FIG. 4 is an explanatory diagram of an example of image data. Here, as a specific example, image data as shown in FIG. 4 is input,
It is assumed that the image data is stored in the image data storage unit 1. In the specific example shown in FIG. 4, the first line reads “Aiue Okaki Kukeko”,
In the second line, “abcdefghijkklmn” is written horizontally, and in the third line, “Nanunnene no hifuheho” is written horizontally.
This image data is displayed to the operator by the display unit 9.

FIG. 5 is an explanatory diagram of line position coordinates and character position coordinates obtained by the character line detection unit 2 and the character cutout unit 3. The character line detector 2 scans such image data in the horizontal direction. Then, in the configuration shown in FIG. 2, the number of black pixels is counted by the horizontal black pixel number counting unit 11 to create a projection, and the line spacing determination unit 12 determines the line spacing. Then, the coordinate y shown by the broken line in FIG.
_{11 to} y ₃₂ are detected. This coordinate becomes the y coordinate of each character and is stored in the character coordinate storage unit 4.

The character slicing unit 3 scans each character line detected by the line space determining unit 12 in the vertical direction and counts black pixels to create a projection. In the example shown in FIG. 3, the line segment coordinates of the first line in the character segmenting unit 3 are y ₁₁ , y _12.
Is obtained, the black pixel is counted in the vertical direction by referring to the band-shaped image data between the coordinates. The character-to-character determination unit 14 cuts out each character based on the created projection to obtain character position coordinates. In the example of the first line shown in FIG. 3, character position coordinates x _{11 to} x _1b are obtained. Here, the coordinates of the left end of each character and the coordinates of the right end of the rightmost character of the line are obtained. Of course, the coordinates of both ends of the leftmost character may be calculated and the coordinate paths of the other characters may be calculated to the right, or the left and right ends of each character may be calculated. Similarly for the second line, character position coordinates x _{21 to} x _2f
However, in the third line, the character position coordinates x _{31 to} x _3b are obtained. The character position coordinates thus obtained are stored in the character coordinate storage unit 4 as the x coordinate of each character. In this way, the character coordinate storage unit 4 stores the character position coordinates as the x-coordinates as coordinate data for each character, and
The row position coordinate is stored as the y coordinate.

The operator designates the start point of the recognition area by the area designating section 5 while referring to the image data displayed on the display section 9. Now, it is assumed that the area specifying unit 5 inputs the coordinates (X0, Y0) of the starting point to the recognition area determining unit 6. Then, the recognition area determination unit 6 causes the character coordinate storage unit 4 to
The coordinate data closest to the coordinate (X0, Y0) in the coordinate data stored in is extracted by the distance calculation. If the extracted coordinates are the coordinates (Xs, Ys), the coordinates are output to the recognition area storage unit 7 as the coordinates of the starting point instead of the coordinates (X0, Y0). For example, when a point near the upper left corner of the character “A” is designated as the start point in FIG. 4, the coordinates (x ₁₁ , y ₁₁ ) are extracted as the coordinates of the start point and stored in the recognition area storage unit 7.

The operator operates the area designating unit 5 while confirming the recognition target area on the display unit 9 in order to specify the other end coordinate of the recognition target area. FIG. 6 is an explanatory diagram of a specific example of rectangular frame display in the first embodiment of the present invention. When the operator is operating the area designation unit 5,
The coordinate area determination unit 6 outputs the coordinates (Xi, Yi) currently designated by the area designation unit 5 at regular time intervals. The recognition area determination unit 6 receives the coordinates (Xi, Y
While i) is instructed, the coordinate data closest to the coordinates (Xi, Yi) is extracted from the coordinate data stored in the character coordinate storage unit 4 by the distance calculation. If the extracted coordinates are the coordinates (Xe, Ye), the coordinates are output to the recognition area storage unit 7 as the coordinates of the end point. The area frame display unit 8 generates an area frame based on the coordinates of the start point and the end point output to the recognition area storage unit 7, and causes the display unit 9 to display the area frame. For example, in FIG. 6, when the area designating unit 5 is moved near the lower right corner of the character “i” or the character “f”, the rectangle indicated by the broken line is sequentially displayed as the area frame along with the movement.

In this way, in the area the operator is going to instruct, the coordinates of the start point (Xs, Ys) and the coordinates of the end point (Xe, Ys) are entered until the final coordinates are input.
The rectangular frame display unit 8 is a rectangular area having e) as diagonal coordinates.
Is displayed on the display unit 9. Finally, for example, when the lower right corner of the character "HI" is designated as the end point, the recognition target area shown by the solid line in FIG. 6 is set. When the final coordinates are designated, the coordinates (Xs, Ys) and the coordinates (Xe, Ye) stored in the recognition area storage unit 7 at that time are
The coordinates are the coordinates indicating the finally specified recognition target area. Then, the recognition unit 18 refers to the coordinate data indicating the recognition target area stored in the recognition area storage unit 7 and performs character recognition on the image data in the area.

When the coordinates designated by the region designating unit 5 are outside the character image, the recognition region determining unit 6 does not change the coordinate data and outputs the designated coordinates as they are to the recognition region storing unit 7. It can also be configured to do so. Further, a modifier key is provided in the area designating unit 5, and the recognition area determining unit 6
Alternatively, when the depression of the modifier key is detected, the designated coordinates may be directly output to the recognition area storage unit 7. As a modifier key, for example, "Shi
It may be an ft "key, a" Control "key, or the like.

The recognition target area is not limited to a rectangle but may be a polygon. In that case, the number of coordinates may be designated according to the shape. Further, the number of recognition target areas to be designated is not limited to one, and a plurality of recognition target areas can be designated.

Next, a second embodiment of the character recognition device of the present invention will be described. FIG. 7 is an explanatory diagram of an example of the case where the area frame overlaps the character in the first embodiment of the present invention. In the above-described first embodiment, the coordinates of the start point and the end point designated by the operator are changed according to the coordinate data stored in the character coordinate storage unit 4 so that they do not overlap with the characters. However, the area frame of the recognition target area set in this way may overlap a character other than near the coordinates of the start point and the coordinates of the end point. Figure 7
Indicates such a case. FIG. 7 shows the area frame when the vicinity of the upper left of the character “b” is specified as the start point and the vicinity of the lower right of the character “I” is specified as the end point. As can be seen from FIG. 7, the left side of the area frame intersects with the character “na”, and the right side intersects with the character “l” and the character “hi”.

In the example shown in FIG. 7, the character spacing differs between the alphabetic characters and the hiragana characters. Therefore, when the recognition target area as shown in FIG. 7 is designated, the conventional recognition target area setting method is used. In order to set a rectangular area that does not intersect a character, it is necessary to change the area for several characters, and the desired recognition target area cannot be set. In order to solve such a problem, the second embodiment of the present invention described below shows an example in which the recognition target area is modified so as to eliminate such an overlap between the area frame and the character image. There is.

FIG. 8 is a block diagram showing the second embodiment of the character recognition device of the present invention. In the figure, the same parts as those in FIG. Reference numeral 31 is an area frame contact detection unit, and 32 is an area frame correction unit. The area frame contact detection unit 31 generates the area frame data of the recognition target area from the coordinates of the start point and the end point of the recognition target area output from the recognition area determination unit 6, and the generated area frame data and the image. It is determined whether or not the character image in the data contacts. For this determination, for example, line segment image data of the area frame is generated,
It can be executed by performing a logical operation with the image data stored in the image data storage unit 1. Alternatively, when the left and right coordinates of each character are detected by the character cutout unit 3 and stored in the character coordinate storage unit 4, the determination is made only by calculating the intersection between the line segment of the area frame and the area where each character exists. can do. When the area frame and the character are in contact with each other, the area frame correction unit 32 expands the area frame by only the part of the character in contact. As a result, all the touching characters are included in the recognition target area. At this time, if the coordinate data stored in the character coordinate storage unit 4 is used, high speed processing is possible.

FIG. 9 is a flow chart showing an example of the operation in the second embodiment of the character recognition device of the present invention.
The operation of the second embodiment is almost the same as that of the first embodiment described above, and therefore, in FIG. 9, the portions overlapping with the flowchart shown in FIG. 3 are omitted except S24 and S25. Further, S24 and S25 are shown by broken lines in FIG. 9, and the contents are the same as in FIG.

Similar to the first embodiment, when the image data is input and stored in the image data storage unit 1, the character line detection unit 2 and the character cutout unit 3 extract the coordinate data and the character coordinate storage unit. Stored in 4. Then, the operator designates the coordinates of the start point and the end point from the area designating unit 5. The coordinates designated by the area designating unit 5 are compared with the coordinate data in the character coordinate storage unit 4 by the recognition area determining unit 6, and the coordinate data closest to the designated coordinates is output from the recognition area determining unit.

When the recognition area determination unit 6 outputs the start point and the end point of the recognition target area, the area frame contact detection unit 31 determines that the area frame determined by the start point and the end point is a character image in S41 of FIG. Detects whether they are in contact. If they are not in contact, the area frame correction unit 32 does not correct the data related to the coordinates of the area frame, and outputs the data to the recognition area storage unit 7 as they are. In this case, the process proceeds to S25.

If there is a contact between the area frame and the character image, the area frame contact detection means 36 detects the coordinates of the contacted character in S42. Then, in S43, it is determined whether the area frame contacting the character is a vertical line segment or a horizontal line segment. This determination result is output to the area frame correction section 37. When the contacted area frame is a horizontal line segment, the area frame correction unit 32 calculates the height of the contacted character in S44 and sets the area frame line of the contacting portion by the height. Extends outward with respect to the frame. That is, the length until contact is stopped is added to the Y coordinate direction in the case of the lower line of the area frame, and is reduced in the case of the upper line, by locally expanding the recognition target area to form a new area frame. create. If the area frame in contact with the character is a vertical line segment, in S45, the width of the contacted character is calculated, and the area frame line of the contacting portion is set to the width of the going frame. To extend outward. That is, when the line on the right side of the area frame is added to the X coordinate direction,
When the line is on the left side, the recognition target area is locally expanded in the decreasing direction to create a new area frame. The newly created area frame is stored in the recognition area storage unit 7.

In S46, it is determined whether or not the processing has been completed for all the contact portions, and if the processing is still left, the processing returns to S43 and is repeated. When the processing of all contact portions is completed, the process proceeds to S25, and the processing after the display of the recognition target area is performed.

FIG. 10 is an explanatory diagram of a specific example of the area frame generated in the second embodiment of the present invention. Now, Fig. 7
Consider the case where the recognition area determination unit 6 outputs the area frame as shown in FIG. As described above, the area frame shown in FIG. 7 is in contact with the character. The area frame contact detection unit 31 detects that the vertical line on the left side of the area frame is in contact with the character “na” and the vertical line on the right side is in contact with the character “l” and the character “hi”. Then, the area frame correction unit 32 is notified. The area frame correction unit 32
First, in the vertical line on the left side of the area frame, the part that is in contact with the character "NA" is expanded to the left until it no longer contacts the character "NA". Further, in the vertical line on the right side of the area frame, the part that is in contact with the character "l" is expanded to the right until it no longer contacts the character "l", and similarly, the part that is in contact with the character "hi" Expand to the right until it no longer touches the letter "hi". By such processing, as shown in FIG.
It is possible to set a recognition target area that does not come into contact with characters.

In the above-mentioned example, the area frame is expanded in order to take in the character touching the area frame as a recognition target, but the area frame is excluded so that the character touching the area frame is excluded from the recognition target. Can be reduced. But,
Even if such a process is not performed, the recognizing unit 10 cannot recognize the missing character, so that the same result can be obtained if the region frame correction process is not performed. A key for instructing not to perform the correction process in the region frame correction unit 32 is arranged in the region designation unit 5, and when the operator presses this key, the region frame correction process is not performed. You can Also, as in the first embodiment,
For example, it is possible to add a function of detecting that the operator has pressed the modifier key and not performing the coordinate changing process by the recognition area determining unit 6. In this case, it is possible not to perform the processing in the area frame correction unit 32.

In the second embodiment described above, the character line detection unit 2 detects the line spacing and outputs the y coordinate of the character. Therefore, the upper and lower frame lines of the area frame output from the recognition area determination unit 6 do not overlap with the characters. Therefore, the area frame contact detection unit 31 can be configured to detect the presence or absence of contact only on the left and right vertical lines. Of course, when the coordinates indicating the area of each character are extracted and stored in the character coordinate storage unit 4, it is necessary to process the upper and lower frame lines.

The above embodiments have been described on the assumption that characters are written horizontally. However, the present invention is not limited to this, and can be applied to a vertically written document.
In that case, the character line detection unit 2 may be configured to scan in the vertical direction to detect line spacing, and the character cutout unit 3 may be configured to scan in the horizontal direction to cut out each character. Further, it is possible to switch between horizontal writing and vertical writing. Further, regardless of horizontal writing or vertical writing, each character may be cut out individually and the coordinates of each character area may be extracted.

[0041]

As is apparent from the above description, according to the present invention, coordinate data relating to a character is detected in advance from the image data stored in the image storage means and stored in the coordinate data storage means. Since the coordinates of the recognition target area specified by the operator are changed, even if the operator specifies the coordinates on the character when specifying the recognition target area, the coordinates that do not overlap the character are calculated and the recognition target area is Is set. Also, when setting this recognition target area,
Since only the coordinate data is processed, the processing can be performed at high speed and the recognition target area designated by the operator can be displayed in real time. for that reason,
It is possible to improve the operability when setting the recognition target area. In addition, when setting the desired recognition target area, if the area frame overlaps with the character, by expanding the area frame for that portion, the desired recognition target area can be efficiently created without causing character loss. The effect is that it can be set.

[Brief description of drawings]

FIG. 1 is a block configuration diagram showing a first embodiment of a character recognition device of the present invention.

FIG. 2 is a block configuration diagram showing an example of a character line detection unit 2 and a character cutout unit 3.

FIG. 3 is a flowchart showing an example of an operation in the first embodiment of the character recognition device of the present invention.

FIG. 4 is an explanatory diagram of an example of image data.

5 is an explanatory diagram of line position coordinates and character position coordinates obtained by the character line detection unit 2 and the character cutout unit 3. FIG.

FIG. 6 is an explanatory diagram of a specific example of rectangular frame display in the first embodiment of the present invention.

FIG. 7 is an explanatory diagram of an example of a case where an area frame overlaps a character according to the first embodiment of this invention.

FIG. 8 is a block configuration diagram showing a second embodiment of the character recognition device of the present invention.

FIG. 9 is a flowchart showing an example of an operation in the second embodiment of the character recognition device of the present invention.

FIG. 10 is an explanatory diagram of a specific example of a region frame generated in the second embodiment of the present invention.

[Explanation of symbols]

1 ... Image data storage unit, 2 ... Character line detection unit, 3 ... Character cutout unit, 4 ... Character coordinate storage unit, 5 ... Area designation unit, 6 ...
Recognition area determination unit, 7 ... Recognition area storage unit, 8 ... Region frame display unit, 9 ... Display unit, 10 ... Recognition unit, 31 ... Region frame contact detection unit, 32 ... Region frame correction unit.

Claims (2)

[Claims] 1. In a character recognition device for recognizing a character image in an image area set based on an instruction from an operator, image storage means for storing image data that can be a target of character recognition, and storage in the image storage means. Coordinate data detecting means for detecting the coordinate data of the lines in the image data and the coordinates of one character or each character component, and coordinate data storing means for storing the coordinate data detected by the coordinate data detecting means. An area specifying means for receiving an area specified by an operator, and a recognition area determining means for calculating coordinates closest to the coordinates specified by the area specifying means in the coordinate data stored in the coordinate data storage means. A character recognition device comprising: 2. An area frame contact detection means for detecting whether or not the area frame of the recognition area specified by the coordinates determined by the recognition area determination means is in contact with a character image, and the area frame contact detection means. 2. The character recognition device according to claim 1, further comprising area frame correction means for expanding or reducing an area frame of a portion in contact with the detected character image so as not to overlap the character.