JPH01241687A - Optical character reader - Google Patents

Abstract

PURPOSE:To prevent a character recognizing rate from being influenced by the accuracy of a format parameter inputted by a user, by correcting character position information for a document, which is later read, with using the character position information of a real segmented result and segmenting the respective characters with the suitable character position information. CONSTITUTION:A character position information correcting part 7 corrects the character position information, which are held by a character position information holding part 3, to the mean value of the character position information of the reading results, according to the average value of comparing difference information received from a character position information comparing part 6 concerning a read N-number of documents. On the other hand, a character image storing part 5 receives the image information of the respective characters segmented by a character image segment part 4 and stores the information successively. A character recognizing part 8 reads the image information of the respective characters stored in the character image storing part 5 and recognizes the character shown by the image information successively and the character codes of the respective characters are generated. Thus, the respective characters can be segmented by the exact character position information.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to an optical character reading device, and particularly to an optical character reading device that recognizes characters on a form that are optically read and generates a character code for the characters. Relating to a reading device.

[Conventional technology]

With conventional optical character reading devices, the user can
By inputting a format parameter indicating character position information in the reading area, the character position in the reading area is designated, and then forms having the character positions in the reading area are read all at once.

[Problem that the invention seeks to solve]

In the conventional optical character reading device described above, when inputting format parameters, the user uses a ruler to measure the position where characters will be written on the actual form, and uses that dimension as character position information in the reading area. I am typing.

For this reason, the user does not necessarily make inputs whose dimensions exactly match, and it is common for format parameters with minute dimensional errors to be input.

In such cases, conventional optical character reading devices cut out characters with minute dimensional errors, resulting in a decrease in the character recognition rate for correctly reading each cut out character. There is a point.

The purpose of the present invention is to use the character position information resulting from actually cutting out to correct the character position information for a form to be read later, and to cut out each character with accurate character position information. To provide an optical character reading device whose character recognition rate is not affected by the accuracy of format parameters input by a user.

[Means for solving problems]

The optical character reading device of the present invention includes: (A>An optical scanner unit that reads an optical signal from a reading area containing a character string on a form and sends out image information of the reading area as an electrical signal; (B) (C) a form image storage unit that stores the image information received from the optical scanner unit as two-dimensional information corresponding to the reading area on the form; (C) a character position in the reading area specified by the user by inputting a format parameter; (D) cutting out image information of each character from the image information stored in the form image storage unit using the character position information held in the character position information storage unit; ,
a character image cutting section that sends character position information of the reading result indicated by black dot bits included in the image information of each character; (E) character position information of the reading result received from the character image cutting section and the character position information; a character position information comparison unit that sends comparison difference information compared with the character position information held in the holding unit; (F) comparison difference information received from the character position information comparison unit for each of the N documents read; (G) receiving image information of each character cut out by the character image cutting section; (F) a character image storage unit that sequentially stores the image information of each character stored in the character image storage unit, sequentially recognizes the characters indicated by the image information, and generates a character code for each character; It is configured with a recognition section.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the optical character reading device of the present invention.

In FIG. 1, an optical scanner unit 1 reads an optical signal from a reading area including a character string on a form, and sends image information of the reading area to a form image storage unit 2 as an electrical signal.

Further, the form image storage section 2 stores the image information received from the optical scanning section 1 as two-dimensional information corresponding to the reading area on the form.

On the other hand, the character position information holding section 3 receives and holds character position information of the reading area specified by the user by inputting format parameters.

Therefore, the character image cutting unit 4 cuts out the image information of each character from the image information stored in the form image storage unit 2 according to the character position information held in the character position information storage unit 3, and extracts the image information of each character from the image information stored in the form image storage unit 2. At the same time, the character position information of the reading result indicated by the coordinate range obtained by projecting the black dot bits of the image information of each character in the X-axis direction and the Y-axis direction is sent to the character position information comparison section 6.

The character position information comparison unit 6 compares the character position information of the reading result received from the character image extraction unit 4 with the character position information held in the character position information holding unit 3 and uses the comparison difference information to the character position information correction unit. Send on 7.

The character position information correction unit 7 calculates the character position information held in the character position information holding unit 3 based on the average value of the comparison difference information received from the character position information comparison unit 6 for each of the N documents read. The character position information of the reading results is corrected to the average value.

On the other hand, the character image storage section 5 receives image information of each character cut out by the character image cutout section 4 and sequentially stores the image information.

The final character recognition section 8 reads out the image information of each character stored in the character image storage section 5, sequentially recognizes the characters indicated by the image information, and generates a character code for each character.

As described above, the optical character reading device of this embodiment corrects the character position information for the form to be read later by using the character position information resulting from the actual extraction, and accurately reads the characters. Each character can be extracted using position information.

〔Effect of the invention〕

As explained above, the optical character reading device of the present invention corrects the character position information for the form to be read later by using the character position information resulting from the actual extraction, and accurately determines the character position. Since each character is extracted using information, the character recognition rate in the optical character reading device of the present invention has the advantage that it is not affected by the accuracy of format parameters input by the user.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the optical character reading device of the present invention. 1... Optical scanner section, 2. Old form image storage section, 3... Character position information holding section, 4.
Old...Character image extraction unit, 5...Character image storage unit, 6...Character position information comparison unit, 7.
...Character position information correction section, 8...Character recognition section.

Claims (1)

[Scope of Claims] (A) an optical scanner unit that reads an optical signal from a reading area including a character string on a form and sends out image information of the reading area as an electrical signal; (B) the optical scanner (C) A form image storage unit that stores image information received from the unit as two-dimensional information corresponding to the reading area on the form; (C) Holds character position information of the reading area specified by the user by inputting format parameters; a character position information holding unit; (D) cutting out image information of each character from the image information stored in the form image storage unit based on the character position information held in the character position information holding unit;
a character image cutting section that sends character position information of the reading result indicated by black dot bits included in the image information of each character; (E) character position information of the reading result received from the character image cutting section and the character position information; a character position information comparison unit that sends comparison difference information compared with the character position information held in the holding unit; (F) comparison difference information received from the character position information comparison unit for each of the N documents read; (G) receiving image information of each character cut out by the character image cutting section; (F) a character image storage unit that sequentially stores the image information of each character stored in the character image storage unit, sequentially recognizes the characters indicated by the image information, and generates a character code for each character; An optical character reading device comprising a recognition unit.