JPH09223181A - Ruled line preparing device, matrix data recognizing device and character recognizing device provided with these devices - Google Patents

Abstract

PROBLEM TO BE SOLVED: To automatically prepare a ruled line only by setting up an area for picture data arranged like a matrix. SOLUTION: The area of picture data arranged like a matrix is set up as an area for preparing a ruled line (step S100), the projection of the picture data in the set area to (x) and (y) axes is generated (S102). Character parts and a blank part in the picture data are recognized based upon the generated projection (S104) and the center point of the recognized blank part is set up as a position for preparing a ruled line (S106). Ruled line data are outputted to the set position to prepare a ruled line (S108). Thus the ruled line can be automatically prepared only by setting up the area for the picture data arranged like a matrix.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a ruled line creating device, a matrix data recognizing device, and a character recognizing device. More specifically, the ruled line creating device creates ruled lines for image data arranged in a matrix in an input image. Matrix data recognition device for recognizing device and image data arranged in matrix in input image as matrix data and device for recognizing image data of character in input image as coded character and consisting of this recognized character The present invention relates to a character recognition device that creates text.

[0002]

2. Description of the Related Art Conventionally, as a character recognition device of this type,
For the image data of the characters surrounded by the ruled line frame, it has been proposed to recognize the image data of the characters as coded characters and replace the ruled lines with delimiters (for example, “,”) to output the text. There is. In this apparatus, by replacing the ruled lines with delimiters, the image data with the ruled line frame can be used as matrix data of coded characters or form data. The matrix data or form data output in this way is read by a spreadsheet device, a graph creating device, or the like, and used for data analysis, tabulation, or graphing.

[0003]

However, the above-mentioned device has a problem that when the image data of characters is arranged in a matrix but there is no ruled line, the image data is not treated as matrix data or form data. . The generation of data breaks in matrix data or form data is performed by detecting ruled lines. Therefore, data breaks cannot be generated for matrix-shaped image data having no ruled lines. Therefore, in order to text-output such matrix-shaped image data without ruled lines as matrix data or form data, it was necessary to manually add ruled lines and then handle them as image data with ruled line frames.

An object of the ruled line creating apparatus of the present invention is to automatically create ruled lines for matrix-shaped image data having no ruled lines by simply designating an area thereof to obtain image data with ruled lines. .

Another object of the matrix data recognition apparatus of the present invention is to recognize matrix-shaped image data without ruled lines as matrix data without creating ruled lines.

The object of the character recognition device of the present invention is to output matrix-shaped or matrix-shaped image data as matrix data or form data with data delimiters and text output.

[0007]

Means for Solving the Problem and Its Action / Effect The ruled line creating apparatus of the present invention is a ruled line creating apparatus for creating ruled lines for image data arranged in a matrix in an input image. Area setting means for setting areas of the image data arranged in a matrix, projection generating means for generating a projection of the image data in the set area in a predetermined direction, and based on the generated projection, The gist is to provide a ruled line creating means for creating a ruled line in the area.

In this ruled line creating apparatus, the area setting means sets the area of the image data arranged in a matrix in the input image, and the projection generating means sets the predetermined direction of the image data in the set area. Generate a projection to.
Then, the ruled line creating means creates a ruled line in the area set based on the generated projection.

According to this ruled line creating apparatus, it is possible to create a ruled line only by designating an area for image data arranged in a matrix.

In such a ruled line forming apparatus of the present invention,
The predetermined direction may be a horizontal direction and / or a vertical direction with respect to the input image.

Further, in the ruled line producing apparatus of the present invention, the projection generating means may be means for taking the integrated value of the pixel values of the image data in the area in the predetermined direction as the projection. . Here, the pixel value means a value 1 or a value 0 in the case of binary data, and in the case of gradation data, a value of the gradation or a value as binary data after being binarized. means. By doing this, it is possible to generate a weighted projection and to create a ruled line at a more appropriate position.

In the ruled line creating apparatus of the present invention, the ruled line creating unit determines the intervals of rows and / or columns in the image data arranged in a matrix in the area based on the projection created by the projection creating unit. It is also possible to provide interval recognition means for recognizing and ruled line data adding means for adding ruled line data to the recognized interval.

In the ruled line creating apparatus according to this aspect, the interval recognizing means included in the ruled line creating means is arranged in rows and / or columns in the image data arranged in a matrix in the area based on the projection generated by the projection generating means. Recognize the spacing. Then, the ruled line data adding means included in the ruled line creating means adds ruled line data to the recognized intervals. With the ruled line creating apparatus of this aspect, it is possible to create a ruled line at a more appropriate position.

In a ruled line creating apparatus including a projection generating apparatus that integrates pixel values of image data to obtain a projection, the ruled line creating means is configured such that the integrated values as the projections created by the projection creating means are consecutive. And an interval recognition means for recognizing a range of the predetermined value or less as a row and / or column interval in the image data arranged in a matrix in the region when the predetermined value is less than or equal to the predetermined value. Ruled line data adding means for adding ruled line data to different intervals may be provided.

In the ruled line creating apparatus according to this aspect, when the interval recognition means provided in the ruled line creating means has the integrated value as the projections integrated by the projection creating means continuously less than or equal to a predetermined value over a predetermined range. A range below the predetermined value is recognized as a row and / or column interval in the image data arranged in a matrix in the area. Then, the ruled line data adding means included in the ruled line creating means adds ruled line data to the recognized intervals. With the ruled line creating apparatus of this aspect, it is possible to create a ruled line at a more appropriate position.

In these ruled line creating apparatuses, when the image data in the area set in the area setting means is character image data, a character frame setting means for setting a character frame for the image data is provided. The projection generating means may be means for generating a projection in the character frame set by the character frame setting means.

In this way, the character portion and the space portion can be made clearer, and the ruled line can be created at a more appropriate position.

A first character recognition device of the present invention is a character recognition device for recognizing image data of a character in an input image as a coded character and creating a text composed of the recognized character. Any one of the ruled line creation apparatus of the present invention described above and a delimiter output unit that outputs a delimiter by a predetermined character or symbol in place of the ruled line in the vertical direction at the time of recognizing the characters of the image data of the area accompanying the ruled line. The point is to prepare.

In the first character recognition device, one of the ruled line creating devices of the present invention described above creates a ruled line for the image data arranged in a matrix, and the delimiter output means thus creates a ruled line. At the time of recognizing the characters of the image data of the area including the ruled line including the defined area, at least the vertical ruled line is replaced with a predetermined character or symbol delimiter.

According to the first character recognition device, matrix-shaped image data without ruled lines can be easily made into matrix data or form data with divisions.

The matrix data recognition device of the present invention is
A matrix data recognition device that recognizes image data arranged in a matrix in an input image as matrix data, and area setting means for setting a region of the image data arranged in the matrix,
Projection generating means for generating a projection of image data in the set area in a predetermined direction, and recognition for recognizing rows and / or columns in the image data arranged in a matrix based on the generated projection The gist is to provide the means.

In this matrix data recognition device, the area setting means sets areas of the image data arranged in a matrix, and the projection generation means projects the image data in the set area in a predetermined direction. To generate. And
The recognition means recognizes the rows and / or columns in the image data arranged in a matrix based on the generated projection.

According to this matrix data recognizing device, it is possible to recognize these data as matrix data only by designating the area for the image data arranged in a matrix.

In the matrix data recognizing device of the present invention, the predetermined direction may be horizontal and vertical with respect to the input image.

In the matrix data recognizing device of the present invention, the projection generating means may be means for taking the integrated value of the pixel values of the image data in the area in the predetermined direction as the projection. Here, the pixel value is
In the case of binary data, it means the value 1 or the value 0, and in the case of gradation data, it means the value of the gradation or the value as the binary data after being binarized. In this way, a weighted projection can be generated and can be recognized more appropriately as matrix data.

In the matrix data recognizing device of the present invention, the recognizing means determines the predetermined value when the integrated value as the projections integrated by the projection creating means is continuously below a predetermined value over a predetermined range. A means for recognizing a range equal to or less than the value as a row and / or column interval in the image data arranged in a matrix in the area may be used. This makes it possible to recognize the matrix data more appropriately.

A second character recognition device of the present invention is a character recognition device for recognizing image data of a character in an input image as a coded character and creating a text composed of the recognized character. When the image data of the area is recognized as matrix data by the matrix data recognition device according to any one of the above-described present invention, the recognized line is recognized when the character of the image data of the region is recognized. The gist is to provide a delimiter output unit that outputs a predetermined character or symbol as a delimiter to and / or a column.

This second character recognizing device, when the image data arranged in a matrix within the region set by any of the matrix data recognizing devices of the present invention is recognized as matrix data, At the time of recognizing the characters of the image data, a predetermined character or symbol is output as a delimiter for the recognized row and / or column.

According to the second character recognition device, matrix-shaped image data without ruled lines can be easily made into matrix data or form data with divisions.

The ruled line creating method of the present invention is a ruled line creating method for creating a ruled line for image data arranged in a matrix in a set area of an input image.
A projection of the image data in the set area in a predetermined direction is generated, and the spacing between rows and / or columns in the image data arranged in a matrix in the area is recognized based on the generated projection. The gist is to create a ruled line by adding ruled line data to the recognized interval.

According to this ruled line creating method, a ruled line can be created only by designating an area for image data arranged in a matrix.

The matrix data recognition method of the present invention is
A matrix data recognition method for recognizing, as matrix data, image data arranged in a matrix in a set region of an input image, and generating a projection of image data in the set region in a predetermined direction. Then
The gist is to recognize the image data in the area as matrix data by recognizing rows and / or columns in the image data arranged in a matrix based on the generated projections.

According to this matrix data recognition method, it is possible to recognize these data as matrix data only by designating the area for the image data arranged in a matrix.

[0034]

Other Embodiments of the Invention The present invention can also take the following other embodiments.

A first aspect as another aspect of the present invention is
It is a storage medium storing a program, and when the program is read by a computer capable of handling an image, the program is to operate the computer as the ruled line creating apparatus of the present invention. In the storage medium of the first aspect, if the program stored in the storage medium is read into a computer capable of handling images, it can be operated as the ruled line creating apparatus of the present invention described above.

A second aspect is a storage medium storing a program, which causes the computer to operate as the matrix data recognition device of the present invention when read into a computer capable of handling an image. The point is that it is a thing. In the storage medium of the second aspect, if the program stored in the storage medium is read into a computer capable of handling images, it can be operated as the matrix data recognition device of the present invention described above.

A third aspect is a storage medium in which a program is stored. When the program is read into a computer capable of handling an image, the computer causes the computer to read the first or second character of the present invention. The gist is that it is operated as a recognition device. In the storage medium of the third aspect, if the program stored in the storage medium is read by a computer capable of handling images, it can be operated as the above-described first or second character recognition device of the present invention. .

[0038]

Next, embodiments of the present invention will be described based on examples. FIG. 1 is a schematic configuration diagram of a character recognition device including a ruled line creation device according to an embodiment of the present invention, and FIG. 2 is a schematic configuration diagram of a computer that realizes the character recognition device. First, for convenience of description, the overall configuration will be described with reference to FIG.

As shown in FIG. 2, this computer 10
Is an arithmetic processing unit 20 connected to the local bus 22, and connects the local bus 22 to a PCI bus 32 which is one of external buses.
To the controller section 40 which is accessed by the CPU 21 and the like of the arithmetic processing section 20 via the PCI bus 32, and a device for controlling various I / O devices is connected to the ISA bus 42 which is a low-speed external bus. The connected I / O unit 60 and the keyboard 7 as a peripheral device
2, a speaker 74, a color CRT 76, a scanner 94 and the like.

The arithmetic processing unit 20 includes a CPU 21 (a Penti manufactured by Intel Corporation in this embodiment) as a central arithmetic processing unit.
umTM), a cache memory 23, a cache controller 24 thereof, and a main memory 25. The PCI bridge 30 is a high-speed PCI bus 3
2 is a controller having a function of controlling the controller 2. CPU
The memory space handled by the CPU 21 is expanded to logical addresses wider than actual physical addresses by various registers prepared inside the CPU 21.

The controller section 40 includes a monitor (color C
(RT) 76, a graphics controller (hereinafter, referred to as CRTC) 44 that controls the display of an image on the RT, and a connected SC.
SCSI controller 4 that controls data transfer with SI equipment
6. The PCI-ISA bridge 48 controls the interface between the PCI bus 32 and the lower ISA bus. The CRTC 44 is 6 compared to the color CRT 76.
It is possible to display 40 × 480 dots in 16 colors. In addition,
Although a character generator that stores a display font, a graphic controller that receives a predetermined command and draws a predetermined graphic, and a video memory that stores a drawn image are mounted on the CRTC 44,
Since these configurations are well known, they are not shown in FIG.

The ISA bus 42 connected via the PCI-ISA bridge 48 is an input / output control bus to which various I / O devices are connected, and includes a DMA controller (hereinafter simply referred to as DMA) 50 and real time. Clock (R
TC) 52, composite I / O port 54, sound I / O5
6, a keyboard interface that controls the keyboard 72 and the two-button mouse 73 (hereinafter referred to as KE
64), an interrupt controller (hereinafter referred to as PIC) 66 that performs interrupt control with priority, a timer 68 that generates various time counts and beeps, and the like. The ISA bus 42 is connected to an ISA slot 62 on which an expansion board can be mounted. In the embodiment, the scanner interface 93 is attached to one of the ISA slots 62,
A scanner 94 is provided via the scanner interface 93.
Is connected.

The composite I / O port 54 is provided with ports for inputting and outputting signals for controlling the floppy disk device 82 and the hard disk 84 in addition to parallel output and serial output. A printer 88 is connected to the parallel input / output via a parallel port 86, and a modem 92 is connected to the serial input / output via a serial port 90. A microphone 96 can be connected to the sound I / O 56 in addition to the speaker 74 described above.

Hard disk 8 of this computer 10
A DOS and various device drivers incorporated in the DOS are stored in the computer 4.
Loads DOS at startup and installs the necessary device driver according to the contents of the file referenced by DOS. As a device driver, there is a printer driver that enables printing to the printer 88 via the composite I / O port 54 and the like.

In the hard disk 84, "WINDOW
An operating system having a GUI "S" is stored ("WINDOWS" is a trademark of Microsoft Corporation), and the computer 10 loads the operating system and then stores the application program in the main memory so that the application program operates on the operating system. Read on top and run.

Next, the character recognition apparatus of this embodiment will be described with reference to FIG. This character recognition device is realized by the hardware and software integrated in the computer 10 described above, and FIG. 1 is a block diagram including the parts realized by the software. This character recognition device has a control unit E
Focusing on CU, keyboard 72 and 2-button mouse 7
3, an input unit IPB that receives the input from the scanner 94 via the operating system, saves the input document or image as a file, and saves the software (application program) necessary to configure the dictionary or the character recognition device A hard disk 84, an image data storage unit GDM for storing image data input by the input unit IPB on a memory, and a ruled line creation unit for creating ruled lines for image data arranged in a matrix based on a predetermined instruction. LPM, a character recognition unit CCM that recognizes image data of characters in an input image as coded characters, a projection generation unit SPM that generates a projection of image data of a specified area on the x-axis and y-axis, a document Equipped with a display unit DPL, etc. that displays images, images, recognition results, etc. on the color CRT 76 .

The ruled line creating section LPM includes an area setting section L1 for setting an area for creating a ruled line in an image and an area setting section L1.
Between the image data generated by the projection generation unit SPM and the projection generation instruction unit L2 for instructing the projection generation unit SPM to generate the projection of the image data in the area set by And a ruled line position setting unit L4 that sets a position for creating a ruled line in the blank determined by the blank determination unit L3.
A ruled line output unit L5 that outputs ruled line data to create a ruled line at a position set by the ruled line position setting unit L4.

The character recognizing section CCM projects an area setting section C1 for setting an area for recognizing characters, and a projection for generating image projections of the image data in the area set by the area setting section C1 on the x-axis and the y-axis. A projection generation instruction unit C2 that gives an instruction to the generation unit SPM, a ruled line determination unit C3 that determines a ruled line from the projection created by the projection generation unit SPM, and image data of a character in an image is recognized and processed as a coded character. A character recognition processing unit C4, a delimiter generation unit C5 that generates a delimiter for a ruled line between characters when the character recognition processing unit C4 recognizes a character, and a recognition result storage unit that temporarily stores a recognition result. And C6.

Then, the projection generator SPM generates an x-axis projection of the set area (projection to the horizontal axis) and an x-axis projection generator S1 of the set area to the y-axis. Y-axis projection generator S2 for generating (projection on vertical axis)
And a projection storage unit S3 for temporarily storing the projection on the x axis and the projection on the y axis.

As described above, each block shown in FIG. 1 includes a part realized by hardware and a part realized by software. For example, the display unit DPL is, as hardware, the CRT shown in FIG.
Although C44 exists, in reality, the software that controls the display including the display of the text of the input character string, the display of the conversion candidates, the display of the guidance of the conversion operation, and the like operates only. Therefore, the display unit DPL and the like in FIG. 1 include both. The software of the character recognition device of the embodiment is stored in the hard disk 84, and when an activation instruction is given from the input unit IPB such as the keyboard 72, the operating system described above reads it into the main memory. Works.

In the character recognition apparatus of the embodiment, data as characters of an image which is normally input from the scanner 94 or the hard disk 84 and stored in the image data storage unit GDM is encoded by the character recognition processing unit C4 of the character recognition unit CCM. It is recognized as a character and the recognition result is output as a document file. Character recognition processing unit C4 of the character recognition unit CCM
The character recognition processing by means of the cutting processing that scans the black dots in the image and cuts out consecutive black dots with a rectangular frame to generate a character frame, and projects the cut-out character frame onto the y-axis (vertical (Line projection to the axis) and to extract lines from this projection, line extraction processing to generate a projection of the clipped character frame to the x axis (projection to the horizontal axis), and this projection and line extraction processing Character extraction processing to extract characters from lines and
A recognition process that outputs a likely coded character as a candidate by analyzing the distribution of black dots in the frame of extracted characters and the way in which black dots are continuous, and determines the likelihood of a recognized character or makes a candidate. And post-recognition processing for correcting

Next, an operation when the image data such as characters arranged in a matrix form without any ruled lines is read in the image input by the character recognition device of the embodiment for performing such recognition processing, that is, in a matrix form. The operation of automatically creating a ruled line in the arranged image data and the operation of recognizing the characters of the image in which the ruled line is created will be described. First, the operation of automatically creating ruled lines in image data arranged in a matrix will be described based on the ruled line creation processing routine of FIG. This routine is executed, for example, when the command of “automatic creation of ruled lines” is selected from the menu bar (not shown) or when the icon indicating “automatic creation of ruled lines” on the function bar is clicked by the two-button mouse 73. . FIG. 4 shows how ruled lines are created by executing this routine.

When this routine is executed, the CPU 21
First, the setting process of a rectangular area (ruled line creation area) W1 for creating a ruled line in the image is performed (step S100).
Specifically, this process is a process of displaying a message box (not shown) in the bottom column of the color CRT 76 so that the operator can set the ruled line creation area W1 in the image, and the operator sets it in the image. 2 on the diagonal of the ruled line creation area W1
By specifying points in the image with the two-button mouse 73 or the keyboard 72, a rectangular area specified by the two specified points is set as the ruled line creation area W1.

When the ruled line creation area W1 is set, a process of generating a projection (x-axis projection) SX of the set area on the x-axis and a projection (y-axis projection) SY of the y-axis (step) is performed. S102). Here, the x-axis projection SX is represented as a histogram created based on the numerical value obtained by counting the black dots of the image data in the ruled line creation region W1 along the y-axis, and the y-axis projection SY is the image data. It is represented as a histogram created based on the numerical values of black dots counted along the x-axis. Therefore, even if the number of projected characters is the same, the black dots in the x-axis direction and the black dot in the y-axis direction differ depending on the characters, and therefore the same projection is not obtained. For example, as in the first to third projections from the left end of the x-axis projection SX in FIG.
Since the number of black dots that are characters but are counted is different, the height of the projection represented as a histogram is different.

When the x-axis projection SX and the y-axis projection SY are generated, a process of recognizing a character portion and a blank portion is performed based on this projection (step S104). Specifically, for example, with respect to the x-axis projection SX, a black dot is 1 m.
The part which is not counted continuously for m or more is judged as a blank part, and the other part is judged as a character part for recognition. For the y-axis projection SY,
The part where two or more consecutive black dots are not counted is determined as a blank part, and the other part is determined as a character part for recognition.

Thus, the x-axis projection SX and the y-axis projection SY
When the blank portion is recognized, the center point of the blank portion is set as ruled line positions kx1 to kx5 and ky1 to ky5 (step S106), and the ruled line data is output to the set position of the image to create the ruled line. Is executed (step S108). In this ruled line creation processing, ruled lines outside the ruled line positions kx1, kx5, ky1, and ky5 at both ends of each axis are deleted. Therefore,
As shown in FIG. 5, the created ruled line is created in the rectangular area surrounded by the ruled line positions kx1, kx5, ky1, and ky5 at both ends of each axis and in the outer edge thereof.

Next, the character recognition processing in the area where the ruled lines are created in this way will be described based on the character recognition processing routine of FIG. This routine is executed, for example, when the "character recognition" command is selected from the menu bar (not shown) or when the icon indicating "character recognition" on the function bar is clicked by the two-button mouse 73.

When this routine is set, the CPU 21
First performs setting processing of an area (character recognition area) W2 in which the character recognition processing is performed in the image (step S11).
0). This processing is step S1 of the ruled line creation processing of FIG.
Similar to the setting processing of the ruled line creation area W1 in 00, processing of displaying a message box (not shown) in the bottom column of the color CRT 76 so that the operator sets the character recognition area W2 in the image, and the operator By designating two diagonal points of the character recognition area W2 to be set in the image with the two-button mouse 73 or the keyboard 72, a rectangular area specified by the designated two points is defined as the character recognition area W2.
And the process of setting as. Although the character recognition area W2 is set in the image here, when a plurality of recognition areas are set in the image in advance, any one of them or two or more is designated as processing. Good. If two or more characters are designated, the character recognition process may be continuously performed in the designated order.

When the character recognition area W2 is set in this way, a process of generating an x-axis projection SX and a y-axis projection SY of the set area is performed (step S112). The x-axis projection SX and the y-axis projection SY here are the steps S in FIG.
This is the same as the projection generation processing described in 102. Subsequently, the peaks in the generated x-axis projection SX and y-axis projection SY are scanned (step S114), and the ruled line is determined based on the detected peak size and shape (step S116). An x-axis projection SX when the image with ruled lines illustrated in FIG. 5 is set as the character recognition area W2 is shown in FIG.
The y-axis projection SY is illustrated in FIG. As described above, since the x-axis projection SX and the y-axis projection SY are represented as a histogram created based on the numerical values of black dots counted along the y-axis or the x-axis, the scale is the same as that of black dots. Then, as shown in FIGS. 7 and 8, the ruled line is represented as it is, and a sharp peak is formed. Therefore, the ruled line can be accurately determined by scanning such a peak and determining the shape thereof.

Next, the image data in the character recognition area W2 is subjected to character recognition processing including the above-described cutout processing, line extraction processing, character extraction processing, recognition processing and post-recognition processing (step S118). ), The vertical ruled line between the characters is replaced with the delimiter “,” and recognized (step S120). For convenience of illustration, the recognition of characters and the replacement of the ruled line with the delimiter "," are described as different steps. However, the replacement of the ruled line with the delimiter "," is performed during the character recognition process. This is done when a vertical ruled line appears on the.

When the characters are recognized in this way, the recognition result, that is, the document file formed by the coded characters is output to the color CRT 76 or the like, and this routine is ended. FIG. 9 shows an example of a document file output as a result of recognizing the ruled image illustrated in FIG. 5 as the character recognition area W2. In the embodiment, as shown in the figure, the ruled line between the characters is replaced by the delimiter “,”, and the blanks are left-justified and output.

According to the character recognition apparatus of the embodiment described above, it is possible to easily create ruled lines for image data arranged in a matrix. Moreover, since the x-axis projection SX and the y-axis projection SY are generated for the image data arranged in a matrix and the ruled lines are created based on these projections, the ruled lines are not created on the character data.

Further, according to the character recognition device of the embodiment, when recognizing an image with ruled lines, the ruled lines between characters can be replaced with the delimiter "," to recognize the characters.
By combining the ruled line creation process and the character recognition process, it is possible to recognize the image data, which has no ruled lines and is arranged in a matrix, as a coded character delimited by the delimiter character “,”. As a result, the image data arranged in a matrix without ruled lines can be made into data that can be easily input by spreadsheet software, graph creation software, etc., and the character resource displayed as an image can be used more effectively. .

In the character recognition apparatus of the embodiment, the x-axis projection SX
For y, the part where black dots are not counted continuously for 1 mm or more is determined as a blank part, and for y-axis projection SY, the part where two or more consecutive black dots are not counted is determined as a blank part. The threshold value for determining blank may be any value. In this case, the operator may set an arbitrary value according to the state of the image, or
The size of one character may be determined from the x-axis projection SX and the y-axis projection SY, and a threshold value for determining a blank portion may be set based on the size of this one character. In the character recognition device of the embodiment, the blank portion is determined for the portion where the black dots are not counted, but the blank portion may be determined for the portion where the black dots less than the predetermined value are counted. . The predetermined value in this case may be set in advance, or may be set based on the size of one character by determining the size of one character from the x-axis projection SX and the y-axis projection SY.

In the character recognition apparatus of the embodiment, the center point of the recognized blank portion is set as the ruled line position, but the x-axis projection S
When setting the ruled line position based on the blank part recognized by X, set the ruled line position based on the blank part recognized by the y-axis projection SY by setting it at a predetermined position viewed from the right end or the left end of the blank part. May be set to a predetermined position viewed from the upper end or the lower end of the blank portion. Further, in the character recognition device of the embodiment, the ruled line is created with the outer edge of the rectangular area surrounded by the ruled line positions (kx1, kx5, ky1, ky5 in the above example) at both ends of each axis as a frame. It is also possible to create ruled lines on the outer edge of the ruled line creation area W1 and not set ruled line positions at both ends.

In the character recognition apparatus of the embodiment, the ruled line creating process and the character recognizing process are individually performed, but a series of processes for performing the character recognizing process after the ruled line creating process may be performed. In this case, the character recognition area W2 can be operated as the ruled line creation area W1, and the processing of steps S112 to S116 in the character recognition processing routine of FIG. 6 can be omitted. This makes it possible to easily recognize characters. In addition, when the ruled line creation processing and the character recognition processing are a series of processing in this way,
Since character recognition and delimiter characters can be performed at the stage of setting the ruled line position, it is not necessary to actually create the ruled line.

In the character recognition apparatus of the embodiment, a histogram created based on the numerical value obtained by counting the black dots of the image data in the ruled line creation area W1 along the y-axis or the x-axis is projected onto the x-axis projection SX or the y-axis. Although SY is used, the simple projection of the image data in the ruled line creation area W1 on the x-axis or the y-axis may be the x-axis projection SX and the y-axis projection SY. That is, 1 on the line along the x-axis or the y-axis.
The projection depending on whether or not there is the above black dot is the x-axis projection SX
And y-axis projection SY. FIG. 10 shows an example of the x-axis projection SX and the y-axis projection SY thus created. As shown, the x-axis projection SX and the y-axis projection SY
Is represented by two states (for example, high level and low level) whether there is a black dot on a line perpendicular to the axis. By doing this, since the projection is generated regardless of the number of black dots on the line, it is possible to determine that there is only a few black dots counted along the y axis, such as the Chinese character "1", and it is determined that the blank part is a blank part. Even a difficult object can be easily determined as a blank portion.

In the character recognition apparatus of the embodiment, the black dots of the image data in the ruled line creation area W1 are counted along the y-axis or the x-axis, but the character recognition processing is performed on the image data in the ruled line creation area W1. Perform the cutting process described in
It is also possible to count the entire cut-out frame as black dots along the y-axis or the x-axis and use the histogram created based on the counted numerical values as the x-axis projection SX or the y-axis projection SY. In this case, the ruled line creation processing routine of FIG. 11 may be executed instead of the ruled line creation processing routine of FIG. In the ruled line creation processing routine of FIG. 11, instead of the projection creation processing of step S102 of the routine of FIG. 3, continuous rectangular dot cutout frame creation processing (step S101) and the created cutout frame are used. A projection generation process (step S103) is performed. FIG. 12 shows an example of the x-axis projection SX and the y-axis projection SY created by this routine. When the projection is generated by the cutout frame in this way, the contrast of the projection becomes clearer, so that it is possible to more appropriately determine the character portion and the blank portion.

Next, a character recognition device which is a second embodiment of the present invention will be described. FIG. 13 is a schematic configuration diagram of the character recognition device in the second embodiment. As shown in the figure, the character recognition device of the second embodiment is different from the character recognition device of the first embodiment (see FIG. 1) in that it includes a matrix data recognition unit MDJ instead of the ruled line creation unit LPM. The character recognizing unit CCM has the same configuration except that it has a matrix recognition result reading unit C7 instead of the ruled line determining unit C3. Therefore, the same components are designated by the same reference numerals and the description thereof is omitted. In addition, the reference numerals attached to the constituent members and functions have the same meaning unless otherwise specified.

The matrix data recognizing unit MDJ provided in the character recognizing device of the second embodiment sets a region for recognizing image data arranged in a matrix as matrix data arranged in rows and columns.
And a projection generation instruction unit M2 for instructing the projection generation unit SPM to generate a projection of the image data in the area set by the area setting unit M1 on the x-axis and the y-axis, and the projection generation unit SPM. A blank determination unit M3 that determines a blank between image data based on the projected image, a matrix setting unit M4 that sets rows and columns in the region based on the blank determined by the blank determination unit M3, and a matrix setting unit M4. The recognition result storage unit M5 temporarily stores the association between the set row and column data and the image data.

The character recognition unit CCM included in the character recognition device of the second embodiment includes the area setting unit C1, the projection generation instruction unit C2, the character recognition processing unit C4, the delimiter character generation unit C5, and the recognition result storage unit C6. In addition, a matrix recognition result reading unit C7 for reading the result of the matrix recognized by the matrix data recognition unit MDJ is provided. The second
The delimiter generation unit C5 of the embodiment generates a delimiter between the characters in different columns when the character recognition processing unit C4 recognizes the character based on the matrix recognition result read by the matrix recognition result reading unit C7.

Next, the matrix data recognition processing in the character recognition apparatus of the second embodiment thus constructed will be described based on the matrix data recognition processing routine of FIG. This routine is executed, for example, when the "Matrix data recognition" command is selected from the menu bar (not shown) or when the icon indicating "Matrix data recognition" on the function bar is clicked by the two-button mouse 73. .

When this routine is executed, the CPU 21
First, a process of setting a rectangular area (matrix data recognition area) W3 to be recognized as matrix data in the image is performed (step S200). Specifically, this process is a process of displaying a message box (not shown) in the bottom column of the color CRT 76 so that the operator sets the matrix data recognition area W3 in the image, and the operator sets it in the image. The two points on the diagonal of the matrix data recognition area W3 to be processed should be the two-button mouse 73 and the keyboard 7.
The process of setting the rectangular area specified by the two points specified by specifying 2 in the image as the matrix data recognition area W3.

When the matrix data recognition area W3 is set, the x-axis projection SX and the y-axis projection SY of the set area are set.
Is generated (step S202), and a process of recognizing a character portion and a blank portion in the x-axis direction and the y-axis direction based on the generated x-axis projection SX and y-axis projection SY is performed (step S204). ). Then, rows and columns are set for the character portion of the image data in the matrix data recognition area W3 based on the recognized character portion and the blank portion in the x-axis direction and the y-axis direction (step S20).
6). This state is illustrated in FIG. As shown
The columns c1 to c4 are set in order from the left for the character portion based on the x-axis projection SX, and the rows r1 to r4 are set in order from the top for the character portion based on the y-axis projection SY.

When the row and column are set, the relation between the position of the image data and the set row and column data is stored in the recognition result storage section M5 as a recognition result, and this routine ends.

Next, the character recognition processing in the area thus recognized as matrix data will be described with reference to FIG.
The character recognition processing routine will be described. This routine is executed, for example, when the "character recognition" command is selected from the menu bar (not shown) or when the icon indicating "character recognition" on the function bar is clicked by the two-button mouse 73.

When this routine is set, the CPU 21
Is the same processing as steps S110 and S112 of the character recognition processing routine of FIG. 6, setting processing of an area (character recognition area) W2 for performing character recognition processing in an image (step S210) and x of the set area. Processing for generating the axial projection SX and the y-axis projection SY is performed (step S2).
12). Subsequently, a process of reading the result recognized by the matrix data recognition process from the recognition result storage unit M5 is performed (step S214). Then, the character recognition area W2
The above-mentioned cut-out processing, line extraction processing,
While performing character recognition processing including character extraction processing, recognition processing, post-recognition processing, and the like (step S216),
A delimiter "," is added between characters in different columns (step S218). For convenience of illustration, the recognition of characters and the addition of the delimiter “,” are described as different steps. However, the addition of the delimiter “,” means that characters in different columns are not recognized during the character recognition process. It is done every time it appears.

When the characters are recognized in this way, the recognition result, that is, the document file formed by the coded characters is output to the color CRT 76 or the like, and this routine is ended. Note that the document file output in this manner has a delimiter "," added between characters in different columns, and blanks are padded to the left and output, as in the document file illustrated in FIG.

According to the character recognition apparatus of the second embodiment described above, it is possible to recognize the image data arranged in a matrix as matrix data having the concept of rows and columns. Further, according to the character recognition device of the second embodiment, it is possible to recognize image data arranged in a matrix without ruled lines as coded characters delimited by delimiters ",". As a result, the image data arranged in a matrix without ruled lines can be made into data that can be easily input by spreadsheet software, graph creation software, etc., and the character resource displayed as an image can be used more effectively. .

In the character recognition device of the second embodiment, the matrix data recognition process and the character recognition process are performed separately, but a series of processes for performing the character recognition process subsequent to the matrix data recognition process. May be In this case, the character recognition area W2 may be operated as the matrix data recognition area W3. In this case, the processing of steps S210 to S214 in the character recognition processing routine of FIG. 16 can be omitted. This makes it possible to easily recognize the image data arranged in a matrix as a character including a delimiter.

In the character recognition device of the second embodiment, the first
Similar to the character recognition apparatus of the embodiment, the histograms created based on the numerical values obtained by counting the black dots of the image data along the y-axis or the x-axis are the x-axis projection SX and the y-axis projection SY. May be simply projected onto the x-axis or the y-axis as the x-axis projection SX and the y-axis projection SY, or the cutout processing described in the character recognition processing may be performed on the image data, and the entire cutout frame may be displayed. The black dots may be counted along the y-axis or the x-axis, and the histogram created based on the counted values may be the x-axis projection SX or the y-axis projection SY.

As described above, the character recognition devices of the first and second embodiments include the part realized by the hardware and the part realized by the software, and the hardware of the software is The non-dependent portion is stored in the hard disk 84 as application software and read as needed, but the medium to be stored is not limited to the hard disk 84, but may be a flexible disk, a magneto-optical disk, or a CDRO.
Any storage medium such as M can be used as long as it is a storage medium that can be used for input / output devices connectable to the computer 10. In this case, the computer 10 directly stores the application software stored in the storage medium (C
It cannot be read from a DROM, a flexible disk, or the like) and can be read only after being installed in the hard disk 84 by a predetermined method.

The character recognition devices of the first and second embodiments have been described as operating on the operation system "WINDOWS", but are assumed to operate as a character recognition device by being read by another operation system. Good.

Although the embodiments of the present invention have been described above, the present invention is not limited to such embodiments. For example, the ruled line creating device and the matrix data recognizing device are not incorporated in the character recognizing device, but are independent. Configuration, a ruled line creation device or matrix data recognition device installed in a device other than a character recognition device that handles image data, a configuration applied to a ruled line creation device that creates ruled lines for image data other than characters, other than characters Needless to say, the present invention can be implemented in various forms without departing from the scope of the present invention, such as a configuration applied to a matrix data recognition device for recognizing image data as matrix data.

[Brief description of drawings]

FIG. 1 is a schematic configuration diagram of a character recognition device including a ruled line creation device that is an embodiment of the present invention.

FIG. 2 is a schematic configuration diagram of a computer in which the character recognition device of the embodiment is realized.

FIG. 3 is a flowchart illustrating a ruled line creation processing routine executed by a CPU 21 of the embodiment.

FIG. 4 is an explanatory diagram illustrating a state when an x-axis projection SX and a y-axis projection SY of image data arranged in a matrix are generated.

FIG. 5 is an explanatory diagram illustrating a state in which a ruled line is created in image data arranged in a matrix.

FIG. 6 is a flowchart illustrating a character recognition processing routine executed by the CPU 21 of the embodiment.

FIG. 7 is an explanatory diagram showing an example of an x-axis projection SX when an image with ruled lines is set as a character recognition area W2.

FIG. 8 is an explanatory diagram showing an example of a y-axis projection SY when an image with ruled lines is set as a character recognition area W2.

FIG. 9 is an explanatory diagram showing an example of a document file output as a result of recognizing an image with a ruled line as a character recognition area W2.

FIG. 10 is an explanatory diagram showing a modified example of the x-axis projection SX and the y-axis projection SY.

FIG. 11 is a flowchart illustrating a part of a modified example of a ruled line creation processing routine.

FIG. 12 is an explanatory diagram showing a modified example of the x-axis projection SX and the y-axis projection SY.

FIG. 13 is a schematic configuration diagram of a character recognition device in a second embodiment.

FIG. 14 is a flowchart illustrating a matrix data recognition processing routine executed by the CPU 21 of the second embodiment.

FIG. 15: x of image data arranged in a matrix
It is explanatory drawing which illustrates the state when the axial projection SX and the y-axis projection SY are produced | generated.

FIG. 16 is a flowchart illustrating a character recognition processing routine executed by the CPU 21 of the second embodiment.

[Explanation of symbols]

 10 ... Computer 20 ... Arithmetic processing unit 21 ... CPU 22 ... Local bus 23 ... Cache memory 24 ... Cache controller 25 ... Main memory 30 ... PCI bridge 32 ... PCI bus 40 ... Controller unit 42 ... ISA bus 44 ... CRTC 46 ... SCSI controller 48 ... ISA bridge 54 ... Composite I / O port 56 ... Sound I / O 60 ... I / O section 62 ... ISA slot 68 ... Timer 72 ... Keyboard 74 ... Speaker 76 ... Color CRT 82 ... Floppy disk device 84 ... Hard disk 86 ... Parallel port 88 ... Printer 90 ... Serial port 92 ... Modem 93 ... Scanner interface 94 ... Scanner 96 ... Microphone CCM ... Character recognition unit C1 ... Area setting unit C2 ... Projection generation instruction unit C3 ... Ruled line determination unit C4 ... Character recognition processing section C5 ... Character generation section C6 ... Recognition result storage section C7 ... Matrix recognition result reading section DPL ... Display section ECU ... Control section GDM ... Image data storage section IPB ... Input section LPM ... Ruled line creation section L1 ... Area Setting section L2 ... Projection generation instruction section L3 ... Blank determination section L4 ... Ruled line position setting section L5 ... Ruled line output section MDJ ... Matrix data recognition section M1 ... Area setting section M2 ... Projection generation instruction section M3 ... Blank determination section M4 ... Matrix setting Part M5 ... Recognition result storage unit SPM ... Projection generation unit S1 ... x-axis projection generation unit S2 ... Y-axis projection generation unit S3 ... Projection storage unit

Claims (14)

[Claims] 1. A ruled line creating device for creating a ruled line for image data arranged in a matrix in an input image, comprising area setting means for setting an area of the image data arranged in the matrix. A ruled line creating apparatus comprising: a projection creating unit that creates a projection of image data in the set area in a predetermined direction; and a ruled line creating unit that creates a ruled line in the area based on the created projection. 2. The ruled line creating apparatus according to claim 1, wherein the predetermined direction is a horizontal direction and / or a vertical direction with respect to the input image. 3. The ruled line creating apparatus according to claim 1, wherein the projection generation means is a means for setting the integrated value of pixel values of image data in the area in the predetermined direction as the projection. 4. The ruled line creating apparatus according to claim 1, wherein the ruled line creating unit is an image arranged in a matrix in the region based on the projection created by the projection creating unit. A ruled line creation apparatus including a space recognition unit that recognizes a space between rows and / or columns in data and a ruled line data adding unit that adds ruled line data to the recognized space. 5. The ruled line creating apparatus according to claim 3, wherein the ruled line creating unit has a predetermined value continuously over a predetermined range of integrated values as projections generated by the projection generating unit. In the following cases, an interval recognition means for recognizing a range equal to or less than the predetermined value as an interval between rows and / or columns in image data arranged in a matrix in the area, and ruled line data is added to the recognized interval. A ruled line creation device including ruled line data adding means. 6. The ruled line creating apparatus according to claim 1, wherein when the image data in the area set by the area setting unit is image data of a character, a character is added to the image data. A ruled line creation device comprising character frame setting means for setting a frame, wherein the projection generation means is means for generating a projection by the character frame set by the character frame setting means. 7. A character recognition device for recognizing image data of a character in an input image as a coded character and creating a text composed of the recognized character, according to any one of claims 1 to 6. A character recognition device comprising: a ruled line creating device; and a delimiter output means for outputting a delimiter by a predetermined character or symbol in place of a ruled line in the vertical direction when recognizing a character in image data in an area accompanied by a ruled line. 8. A matrix data recognition device for recognizing, as matrix data, image data arranged in a matrix in an input image, and area setting means for setting an area of the image data arranged in a matrix. A projection generating unit that generates a projection of image data in the set area in a predetermined direction, and recognizes rows and / or columns in the image data arranged in a matrix based on the generated projection. Matrix data recognition means including a recognition means. 9. The matrix data recognition device according to claim 7, wherein the predetermined direction is a horizontal direction and a vertical direction with respect to the input image. 10. The ruled line creating apparatus according to claim 8, wherein the projection generation unit is a unit that sets an integrated value of pixel values of image data in the area in the predetermined direction as the projection. 11. The recognizing means, when the integrated value as a projection generated by the projection generating means is continuously less than or equal to a predetermined value over a predetermined range, the range less than or equal to the predetermined value is defined as the area. 11. The matrix data recognition device according to claim 10, which is a means for recognizing the intervals of rows and / or columns in the image data arranged in a matrix. 12. A character recognition device for recognizing image data of a character in an input image as a coded character and creating a text composed of the recognized character, wherein: A matrix data recognition device, and when the matrix data recognition device recognizes the image data of the area as matrix data, a predetermined number is specified for the recognized row and / or column when recognizing the characters of the image data of the area. A character recognition device comprising: a delimiter output unit that outputs the character or symbol as a delimiter. 13. A ruled line creating method for creating a ruled line for image data arranged in a matrix in a set region of an input image, wherein a predetermined direction of the image data in the set region is set. On the basis of the generated projection, recognize the row and / or column intervals in the image data arranged in a matrix in the region based on the generated projection, and add ruled line data to the recognized intervals. A ruled line creation method that creates a ruled line by doing so. 14. A matrix data recognition method for recognizing, as matrix data, image data arranged in a matrix in a set area of an input image, the predetermined direction of the image data in the set area. A matrix for recognizing the image data in the area as matrix data by recognizing the rows and / or columns in the image data arranged in the matrix based on the generated projection. Data recognition method.