JPH10334184A - Ruled line erasing method and device, table processing method and device, character recognition method and device and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform accurate ruled line recognition and character recognition even in the case that a character and a ruled line are in contact or the like by extracting a rectangle for which the black runs of more than a prescribed threshold length are extracted from binary images and integrated as the ruled line and converting the black picture elements of the binary images for constituting the ruled line to white picture elements. SOLUTION: The original of a document or the like is read in a binary image input part 1 and is stored in a binary image memory 2. A black run extraction part 3 extracts the black runs of a length more than a fixed value from the binary image memory 2 and stores them in a black run memory 4. A ruled line extraction part 5 integrates the black runs present within a predetermined distance together, extracts them as the ruled lines and stores them in a ruled line memory 6. A ruled line thickness detection part 7 reads the ruled lines from the ruled line memory 6 and calculates the thickness of the ruled lines and a frame extraction part 8 combines the ruled lines, extracts them as a frame area and stores them in a frame memory 9. Then, a ruled line erasing part 10 refers to the binary image memory 2 and the ruled line memory 6, replaces the black picture elements equivalent to the ruled line from the binary images with the white picture elements and erases the ruled line.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a method, an apparatus, a table processing method, an apparatus, a character recognition method, an apparatus, and a recording medium for eliminating ruled lines in a document image containing ruled lines and underlines.

[0002]

2. Description of the Related Art In general, when processing a document in a document recognition apparatus, a document image is often classified into a character area, a table area, a figure, and other areas, and processing suitable for each area is often performed. Among them, as a method of processing a table area, there is a table processing method of recognizing a ruled line constituting a table and extracting a character surrounded by the ruled line before recognizing the character (Japanese Patent Laid-Open No.
1729884).

[0003]

However, according to the conventional method, when a character is in contact with a ruled line or when a character is written through a ruled line frame, a part of the character is erased. Therefore, there is a problem that character recognition cannot be performed accurately.

An object of the present invention is to provide a ruled line erasing device capable of performing accurate ruled line recognition and character recognition even when a character is in contact with a ruled line or when a character is written through a ruled line frame. A method, a device, a table processing method, a device, a character recognition method, a device, and a recording medium are provided.

[0005]

In order to achieve the above object, according to the present invention, a black run having a length equal to or more than a predetermined threshold value is extracted from a binary image, and the black run is extracted. By integrating black runs within a predetermined distance, a rectangle including all the black runs is extracted as a ruled line,
The ruled line is erased by converting black pixels of the binary image forming the ruled line into white pixels.

According to the second aspect of the present invention, a means for inputting a document image as a binary image, a means for extracting a black run having a length equal to or more than a predetermined threshold from the binary image, Means for extracting a rectangle including all the black runs as ruled lines by integrating black runs within a predetermined distance from each other, and converting black pixels of the binary image forming the ruled lines into white pixels. And means for erasing the ruled line.

According to the third aspect of the present invention, a black run having a length equal to or longer than a predetermined threshold is extracted from the binary image, and black runs within a predetermined distance of the extracted black run are integrated. And extracting a rectangle including all the black runs as ruled lines, combining the ruled lines to extract a frame, and converting the black pixels of the binary image forming the ruled lines into white pixels to erase the ruled lines. Then, when the ruled line is deleted, if an interval between white pixels having the same width as the thickness of the ruled line is generated, the interval between the white pixels is converted into a black pixel, and a frame area wider than the extracted frame area is converted. It is characterized by setting and extracting characters from within the set frame.

According to a fourth aspect of the present invention, a means for inputting a document image as a binary image, a means for extracting a black run having a length equal to or more than a predetermined threshold from the binary image, Means for extracting a rectangle including all black runs as ruled lines by integrating black runs within a predetermined distance with respect to each other, means for extracting a frame by combining the ruled lines, and means for erasing the ruled lines. Means for restoring the character image by converting the interval between white pixels to black pixels when an interval between white pixels having the same width as the thickness of the rule line is generated in the character image when the ruled line is deleted. And means for setting a frame area wider than the extracted frame area, and means for extracting characters from within the set frame.

According to a fifth aspect of the present invention, a character extracted by the method of the third aspect is recognized.

According to the present invention, a means for inputting a document image as a binary image, a means for extracting a black run having a length equal to or more than a predetermined threshold from the binary image, Means for extracting a rectangle including all black runs as ruled lines by integrating black runs within a predetermined distance with respect to each other, means for extracting a frame by combining the ruled lines, and means for erasing the ruled lines. Means for restoring the character image by converting the interval between white pixels to black pixels when an interval between white pixels having the same width as the thickness of the rule line is generated in the character image when the ruled line is deleted. Means for setting a frame area wider than the extracted frame area, means for extracting a character from the set frame, and means for recognizing the extracted character.

According to the present invention, a function of inputting a document image as a binary image, a function of extracting a black run having a length equal to or more than a predetermined threshold from the binary image, and a function of extracting the black run A function of extracting a rectangle including all black runs as ruled lines by integrating black runs within a predetermined distance, a function of extracting a frame by combining the ruled lines, and a function of deleting the ruled line. A function of restoring a character image by converting the interval between white pixels to a black pixel when an interval between white pixels having the same width as the thickness of the ruled line is generated in the character image when the ruled line is deleted. A program for causing a computer to realize a function of setting a frame area wider than the extracted frame area, a function of extracting characters from the set frame, and a function of recognizing the extracted characters. Con Is characterized in that Yuta is readable recording medium.

[0012]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be specifically described below with reference to the drawings. <Embodiment 1> FIG. 1 shows the structure of Embodiment 1 of the present invention.
FIG. 2 shows a processing flowchart of the first embodiment. Less than,
The processing operation of the present invention will be described with reference to FIGS.

By a binary image input unit 1 such as a scanner,
An original such as a document or a form is read as a binary image and stored in the binary image memory 2 (step 101). The black run extraction unit 2 extracts a black run having a length equal to or more than a predetermined value from the binary image memory 2 and stores it in the black run memory 4 (step 102). Next, the ruled line extraction unit 5 integrates black runs within a predetermined distance from the black run memory 4 as constituting the same ruled line, and defines a rectangle including all the black runs as a ruled line. It is extracted and stored in the ruled line memory 6 (step 103).

Subsequently, the ruled line thickness detecting section 7 reads the ruled line from the ruled line memory 6, counts the number of black pixels (black) from the black runs constituting the ruled line, and calculates the ruled line length (le).
ngth) and the ruled line thickness (thickne)
ss) (step 104).

Thickness = black / length The above processing is performed not only in the main scanning direction but also in the sub-scanning direction.

The frame extracting section 8 combines ruled lines in the main scanning direction and the sub-scanning direction, extracts an area surrounded by four sides as a frame area, and stores the frame area in the frame memory 9 (step 105). Next, the ruled line erasing unit 10 erases the ruled line by referring to the binary image memory 2 and the ruled line memory 6 and replacing black pixels corresponding to the ruled lines with white pixels from the binary image (step 106).

Next, the character pixel interpolating unit 11 refers to the binary image memory 2 and finds any black pixels which are in contact with the upper and lower parts of the area erased as the ruled line and whose intervals are within a certain value. For example, black pixel interpolation is performed (step 107).

This will be described with reference to the example of FIG. (A) shows an example of an image after ruled line deletion. The black pixel position (Xus, Xue) in contact with the upper side of the erased ruled line area and the black pixel position (Xds, Xde) in contact with the lower side are | Xus−Xde | <Th | Xue−Xds | <Th If either of the conditions is satisfied (Th is a predetermined threshold), the width is (Min (Xus, Xds) -Max (Xu
e, Xde)), and replace the corresponding coordinates in a rectangle with a height of “thickness” with black pixels. (B) shows an image after interpolation. As a result, it is possible to restore the character divided by the ruled line erasure.

The character extracting unit 12 sets an area wider than the frame area extracted in step 105 so that the character fits within the frame area. A rectangle is extracted, characters in the frame are extracted, and stored in the character image memory 13 (step 108). Then, the character recognition unit 14 recognizes the extracted character and obtains a text output (step 109).

<Embodiment 2> The present invention is not limited to the above-described embodiment, but can be realized by software. When the present invention is realized by software,
As shown in FIG. 4, a general-purpose processing device including a CPU, a ROM, a RAM, a display device, a hard disk, a keyboard, a CD-ROM drive, and a scanner is prepared.
On a computer-readable recording medium such as a D-ROM, a program for realizing the table processing and character recognition processing functions of the present invention is recorded. Images such as documents input from a scanner or the like are temporarily stored on a hard disk or the like. Then, when the program is started,
The temporarily stored image data is read, a table process and a character recognition process are executed, and the results are output to a display, a printer, or the like.

[0021]

As described above, according to the present invention, even if a character is in contact with a ruled line or a character is written through a ruled line frame, the character can be accurately recognized. It becomes possible.

[Brief description of the drawings]

FIG. 1 shows a configuration of a first exemplary embodiment of the present invention.

FIG. 2 shows a processing flowchart of the first embodiment.

FIG. 3A shows an example of an image after ruled line erasure, and FIG.
Indicates an image after black pixel interpolation.

FIG. 4 shows a configuration of a second exemplary embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 1 binary image input unit 2 binary image memory 3 black run extraction unit 4 black run memory 5 ruled line extraction unit 6 ruled line memory 7 ruled line thickness detection unit 8 frame extraction unit 9 frame memory 10 ruled line elimination unit 11 character pixel interpolation unit 12 Character extraction unit 13 Character image memory 14 Character recognition unit

Claims (7)

[Claims] 1. A black run having a length equal to or greater than a predetermined threshold is extracted from a binary image, and black runs within a predetermined distance of the extracted black runs are integrated to include all the black runs. A rectangle to be extracted is extracted as a ruled line, and the ruled line is erased by converting black pixels of the binary image forming the ruled line into white pixels. 2. A means for inputting a document image as a binary image, a means for extracting a black run having a length equal to or more than a predetermined threshold value from the binary image, and Means for extracting a rectangle including all the black runs as ruled lines by integrating certain black runs, and converting the black pixels of the binary image forming the ruled lines to white pixels to convert the ruled lines into white pixels A ruled line erasing apparatus comprising means for erasing. 3. A black run having a length equal to or greater than a predetermined threshold is extracted from a binary image, and black runs within a predetermined distance of the extracted black runs are integrated to include all the black runs. A rectangle to be extracted is extracted as a ruled line, a frame is extracted by combining the ruled line, and the ruled line is erased by converting black pixels of the binary image forming the ruled line into white pixels, thereby erasing the ruled line. Then, if an interval between white pixels having the same width as the ruled line thickness is generated, the interval between the white pixels is converted into a black pixel, and a frame area wider than the extracted frame area is set. A table processing method characterized by extracting characters from inside a frame. 4. A means for inputting a document image as a binary image, a means for extracting a black run having a length equal to or greater than a predetermined threshold value from the binary image, and By integrating certain black runs, a means for extracting a rectangle including all the black runs as a ruled line, a means for extracting a frame by combining the ruled lines, a means for deleting the ruled line, and a method for deleting the ruled line When an interval between white pixels having the same width as the ruled line thickness is generated in the character image, a unit for restoring the character image by converting the interval between the white pixels into black pixels, and the extracted frame A table processing apparatus comprising: means for setting a frame area wider than the area; and means for extracting characters from within the set frame. 5. A character recognition method for recognizing a character extracted by the method according to claim 3. 6. A means for inputting a document image as a binary image, a means for extracting a black run having a length equal to or greater than a predetermined threshold value from the binary image, and By integrating certain black runs, a means for extracting a rectangle including all the black runs as a ruled line, a means for extracting a frame by combining the ruled lines, a means for deleting the ruled line, and a method for deleting the ruled line When an interval between white pixels having the same width as the ruled line thickness is generated in the character image, a unit for restoring the character image by converting the interval between the white pixels into black pixels, and the extracted frame A character recognition device comprising: means for setting a frame area wider than the area; means for extracting characters from within the set frame; and means for recognizing the extracted characters. 7. A function for inputting a document image as a binary image, a function for extracting a black run having a length equal to or greater than a predetermined threshold from the binary image, and a function for extracting the black run within a predetermined distance. By integrating certain black runs, a function of extracting a rectangle including all black runs as a ruled line, a function of extracting a frame by combining the ruled lines, a function of deleting the ruled line, and a function of deleting the ruled line When an interval between white pixels having the same width as the ruled line thickness is generated in the character image, the function of restoring the character image by converting the interval between the white pixels to black pixels, and A computer readable recording medium storing a program for causing a computer to implement a function of setting a frame area wider than the area, a function of extracting characters from the set frame, and a function of recognizing the extracted characters. Recording medium.