JPS63167991A - Character recognition device - Google Patents

Abstract

PURPOSE:To perform image processing even when there is a graphic area present at an optional place by retrieving extraction information specified according to the recognition result of a character recognizing means, and outputting a graphic form image when there is a recognition result corresponding to it. CONSTITUTION:A slip P is read by a read part 1 and converted into image data, which is stored in an image buffer 2. A labeling part 3 scans the data to detect a character string image and a graphic image, labels them respectively, and generates a label table wherein those label are related with each other and stores the table in a label storage part 4. An image extraction part 7 retrieves the extraction information read out of a format information storage part 5 among character string image recognized by a character recognition part 6 and when there is a recognition result corresponding to it, the storage part 4 is referred to and graphic images relative to the recognition result corresponding to the extraction information is read out of the buffer 2, and then processed and outputted.

Description

DETAILED DESCRIPTION OF THE INVENTION [Statement of the Invention] (Field of Industrial Application) The present invention relates to a character recognition device that performs character recognition and graphic image processing.

(Prior Art) Character recognition devices that read forms and recognize characters have been widely used in various business processing fields.

In such a character recognition device, a document is read to obtain an overall image, and characters are recognized based on this image.

Therefore, it is possible to perform image processing that outputs an image of a form, and by specifying an image processing area using format information, it is possible to perform image processing of, for example, a graphic image of an IPI form.

However, with such conventional character recognition devices, it is impossible to specify the image processing area using format information, for example, in forms such as general documents where graphics exist anywhere in the document. It was difficult to do so.

(Problems to be Solved by the Invention) The present invention is intended to solve the above conventional problems, and performs image processing of graphic areas even in forms where graphic areas exist anywhere in the document. The purpose of this invention is to provide a character recognition device that can perform the following functions.

[Configuration of the invention (means for solving the problem) In order to achieve the above object, the present invention
! A reading means for obtaining an image of the form, an image storage means for storing the image of the form obtained by the reading means, a character string image and a graphic image by scanning the image storage means, and a character string image detected by scanning the image storage means. and a labeling means for labeling each figure image, and a label table storing a label corresponding to a character string image labeled by this labeling means and a label of a figure image related to this character string image in association with each other. storage means, extraction information specification means for extracting a figure image, character recognition means for character recognition from the character string image stored in the image storage means, and extracting information from the recognition result of the character recognition means. The extracted information specified by the specifying means is searched, and when a recognition result corresponding to this extracted information exists, the image storing means refers to the label table storage means and stores a graphic image related to the recognition result corresponding to the extracted information. and an image extracting means for reading out and outputting the image.

(Operation) The present invention searches for extracted information specified by the extracted information specifying means from the recognition result of the character recognition means, and when there is a recognition result corresponding to this extracted information, extracts it by referring to the label table storage means. Since the configuration is configured so that the graphic image related to the recognition result corresponding to the information is read out from the image storage means and output, image processing of the graphic area can be performed even if the document has a graphic area anywhere in the document. be able to.

(Embodiment) An embodiment of the present invention will be described below in detail with reference to the drawings.6 Fig. 1 is a block diagram showing the configuration of a character recognition device according to an embodiment of the present invention.

In the figure, 1 is a reading unit that optically operates the form to obtain an image of the entire form, 2 is an image buffer that temporarily stores the form image, and 3 is a character string that scans the stored form image. A labeling section 4 detects an image and a figure image and labels the detected character string image and figure image, respectively. A label table storage unit stores labels of related graphic images in association with each other, and 5 is a format information storage unit that stores various information for reading characters written on a form, such as the position of the read character on the II form. , character pitch, and designation of the type of character to be written in the character entry position on the form, as well as extraction information for extracting a predetermined graphic image. Format information including this extracted information is prepared for each type of form and stored in advance. Further, 6 is a character recognition unit that recognizes characters from the character string image stored in the image buffer 2, and 7 is a recognition result of the character string image recognized by the character recognition unit 6, which is stored in the format information storage unit 5. When the extraction information specified by the extraction information is searched and there is a recognition result corresponding to this extraction information, the label table storage unit 4 is referred to and the graphic image related to the recognition result corresponding to the extraction information is stored in the image buffer 2. 8 is an interface section that controls data input/output with the host system; 9 is a control section that controls each section.

In the character recognition device configured as described above, recognition of characters on a form and output of graphic image processing are performed in the following manner.

FIG. 2 is a plan view showing a form to be read. On this form P, horizontally written character strings L1, L2, L3 and a figure G are recorded.

This form P is first read by the reading section 1 and converted into image data. The converted image data is then temporarily stored in the image buffer 2. Thereafter, the labeling unit 3 scans the image data stored in the image buffer 2 to detect character string images and graphic images, and labels these character string images and graphic images respectively. , as shown in FIG. 3, labels ■, ■, ■ are attached to the character strings L1, L2, L3 of the form P shown in FIG.
A label ■ is attached to the figure G. Next, a label table is created by associating the label corresponding to the labeled character string image with the label of the figure image related to this character string image, and the label table is stored in the label table storage unit 4. Store. As shown in FIG. 4, the above-mentioned label table stores labels ■, ■, ■ corresponding to each character string image,
In this case, an attribute label ■, which is a label of a graphic image related to a character string image of label ■, is stored in correspondence.

As shown in FIG. 5, the extracted information stored in the format information storage unit 5 is a keyword "A figure" (A is any alphabetic character), which is a target character or character string.
It is specified that figures within a predetermined distance from "Figure A" are to be targeted.

After that, the image extracting unit 7 uses the character recognition unit 6 to scan the character string image in the image buffer 2 to recognize the characters, and the character recognition result of the character string image recognized by the character recognition unit 6 is The extraction information (keyword K) read from the format information storage unit 5 is searched, and when there is a recognition result (Figure A) corresponding to this extraction information, the label table storage unit 4 is referred to and the extraction information is retrieved. related to the recognition result corresponding to (attribute label exists)
A graphic image is read out from the image buffer 2, processed, and output.

Therefore, in this embodiment, even if it is a form such as a general document in which a figure exists anywhere in the document, the figure image to be subjected to image processing is based on the extraction information, the result of character recognition of the character string image, and the extraction information. It is possible to extract a graphic image by referring to the label table and output the image processing.

[Effects of the Invention] As explained above, the character recognition device of the present invention searches for extracted information specified by the extracted information specifying means from the recognition results of the character recognition means, and there exists a recognition result corresponding to this extracted information. Since the configuration is configured so that the graphic image related to the recognition result corresponding to the extracted information is read from the image storage means and output by referring to the label table storage means, it is possible to read a graphic image related to the recognition result corresponding to the extracted information from the image storage means and output it. Even if there is, it is possible to perform image processing on the graphic area.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing the configuration of a character recognition device according to an embodiment of the present invention, Fig. 2 is a plan view for explaining a vA slip to be read, and Fig. 3 is a label shown in Fig. 2 in the labeling section. FIG. 4 is a diagram for explaining the label table, and FIG. 5 is a diagram for explaining keywords. 1...Reading unit 2...Image buffer 3...
... Labeling section 4 ... Label table storage section 5 ...
...Format information storage unit 6...
・Character recognition section 7...Image extraction section Applicant: Toshiba Corporation Patent attorney Satoshi Suyama - Figure 2, Figure 4, Figure 5L

Claims (1)

[Claims] (1) A reading means that reads a form to obtain an image of the form, an image storage means that stores the image of the form obtained by this reading means, and a character string image and a graphic image are detected by scanning this image storage means. A labeling means for labeling each of the detected character string image and figure image, a label corresponding to the character string image labeled by this labeling means, and a label of the figure image related to this character string image. a label table storage means for storing the images in association with each other; an extraction information specifying means for extracting the graphic image;
a character recognition means for character recognition from a character string image stored in the image storage means; and a character recognition means for searching the extracted information specified by the extraction information specifying means from the recognition result of the character recognition means and corresponding to the extracted information. The present invention is characterized by comprising an image extracting means that refers to the label table storage means when a recognition result exists, reads out from the image storage means a graphic image related to the recognition result corresponding to the extracted information, and outputs it. Character recognition device.