JPH08249421A - Recognizing method for reverse character - Google Patents

Abstract

PURPOSE: To provide a character recognizing method capable of recognizing even a black/white reverse character. CONSTITUTION: An image scanner 21 designates an area to inspect the presence/ absence of a reverse character for binary document image data obtained by applying optical scan to a reading target document, and counts the number of black and white picture elements of a designated area, respectively, and finds a value of black picture element density feature, and when the value of black picture element density feature exceeds a reverse discrimination reference value decided as reference in advance, picture element data in the designated area is reversed, and when the former is smaller than the latter, no reverse processing of the picture element data in the designated area is performed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for recognizing characters in an optical character reader, and more particularly to a method for recognizing black and white reverse printed characters.

[0002]

2. Description of the Related Art A character reading device (OCR) 1 includes an image scanner 21 and a character recognition processor 22 as shown in FIG.
Character reading recognition unit 2 and host computer 3
The character recognition processor 22 receives image data obtained by optically scanning the document to be read by the image scanner 21 as input, and first observes the image data according to the process of the flow shown in FIG. Then, the character pattern data area to be processed as one character is cut out, the cut-out character pattern data is analyzed to extract the characteristic parameter, and this characteristic parameter is converted into each character prepared in advance corresponding to each character in the reading range. The basic function is to recognize a read character by extracting a character with which the characteristic parameter matches by collating with the dictionary of the characteristic parameter to which the character belongs and to output the character classification code assigned to the character as character information. . At this time, the host computer 2
Functions as a man-machine interface for setting the reading conditions of the document to be read and displaying the reading result, and also performs editing and proofreading of the document based on the document information obtained by reading or building a database. To do.

[0003]

In the character recognition method according to the prior art, when there is an area printed with reverse characters as shown in FIG. 2 for the purpose of emphasizing a finding etc. in the document to be read, This is treated as a non-character area similar to figures and photographs, and is not subject to character recognition.
However, headings and the like are often printed in reverse characters for emphasis, and there is also a need to recognize them.

An object of the present invention is to provide a character recognizing method which answers this need and can recognize a reversed character.

[0005]

In the character recognition method according to the present invention, as a preprocessing of character recognition, inversion characters of binarized document image data obtained by optically scanning a document to be read by an image scanner are displayed. Specify the area to be inspected for existence, count the number of black and white pixels in this specified area to obtain the value of the black pixel density feature, and if the value of the black pixel density feature exceeds the inversion discrimination reference value set as a reference in advance. If so, the pixel data of the designated area is inverted, and if the value of the black pixel density feature does not exceed the inversion determination value, the processing of not inverting the designated area pixel data is provided.

[0006]

When a region to be inspected for inversion characters is designated in the binarized document image data input by the image scanner, the value of the black pixel density feature is obtained for this designated region, and the value of the black pixel density feature is used as a reference in advance. If it exceeds the inversion discrimination reference value defined as, the image data corresponding to the non-inverted normal print character is obtained by inverting the binarized pixel data of the specified area, and the value of the black pixel density feature is If the reversal judgment value is not exceeded, the reversal processing of the specified area pixel data is not performed, so the character image data of the characters normally printed remains as it is, and in the character recognition processing in the subsequent stage, all characters are assumed to be non-reversed normal printing To process.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows the flow of processing in one embodiment of the method for recognizing reverse printed characters according to the present invention as shown in FIG. 2, and the method of the present invention will be described with reference to FIG. The configuration of the character reading device that executes the method of the present invention is equivalent to the device of the configuration illustrated in FIG. 3 used in the description of the prior art, and the reference numerals that are given in FIG. 3 when necessary for the following description. To quote.

The character recognition processor 22 of the character reader is
Before reading the binarized document image data scanned and input by the image scanner 21, the image data is temporarily displayed on the display device 31 of the host computer 3 (S2). Then, the reading direction, reading range and other reading conditions at the time of reading the document image data are set through the host computer 3 in consideration of the display contents of the input display screen (S31), but in the method of the present invention, this reading is performed. There is provided a step (S3) of selecting and designating a region for inspecting whether or not characters are reverse printed in the image data of the document input in the condition setting item.

When the reverse printing inspection area is designated, the number of pixels of each black pixel and white pixel is counted in the image data of the character of the designated area, and the value of the black pixel density feature Fb indicating the ratio occupied by the black pixels is calculated. (S4). Figure 2
As seen in the above example, the normal ratio of black pixels in the reverse printed character area is larger than that in the non-reversed printed area. Therefore, when the value Fb of the black pixel density feature for the inversion designated area is obtained, the value of this Fb is selected to exceed the maximum value Fbm of the black pixel density feature empirically obtained for the non-inversion printed character area. It is compared with the set reversal determination reference value th (S5), and if the value of Fb exceeds the determination reference value th, it is possible to determine that the designated area is black-and-white reverse printed area. A process of inverting the data is performed (S6).

On the other hand, when the value Fb of the black pixel density feature for the designated area does not exceed the inversion discrimination reference value th, it is determined that the printing is not inversion, and the inversion processing of the binarized pixel data is not performed. At the stage where the above steps S5 and S6 are completed, the character data of the reverse character is converted into the non-reverse character equivalent, and the character pixel data as the non-reverse character is obtained for all the character areas of the input document image. Character recognition processing is executed by the conventional method to recognize the character (S7).

[0011]

According to the preprocessing method in the character recognition processing of the present invention, even if the document to be read is a document containing black and white reverse printing characters, if the reverse printing area is designated as the reverse inspection area, black pixels will be added to the area. The value of the density feature is obtained, and when this value exceeds the inversion judgment reference value, the binarized pixel data in the designated area is inverted to obtain the character data corresponding to the non-inverted normal print character. Even in a document in which reverse printed characters and non-reversed normal printed characters are mixed, there is no need to separate the reverse printed areas for separate processing, and the efficiency of document reading recognition can be improved.

[Brief description of drawings]

FIG. 1 is a flow chart of a character recognition preprocessing method according to the present invention.

FIG. 2 is a diagram showing an example of light and shade reverse printing characters.

FIG. 3 is a block diagram of a character reading device

FIG. 4 is a basic flow chart of character recognition processing in the character reading device.

[Explanation of symbols]

 1 character reading device 2 character recognition device 21 image scanner 22 character recognition processor 3 host computer

Claims (1)

[Claims] 1. As a pre-processing of character recognition, an area for inspecting the presence or absence of a reverse character is designated in binarized document image data obtained by optically scanning a document to be read by an image scanner, and the designated area is designated. The value of the black pixel density feature is obtained by counting the number of each black and white pixel, and if the value of the black pixel density feature exceeds the inversion determination reference value set as a reference in advance, the pixel data of the designated area is inverted, A method for recognizing a reversed character, characterized in that a process is provided to prevent pixel data in a designated area from being reversed when the value of the black pixel density feature does not exceed the inversion discrimination value.