JPS63103391A - Optical character reader - Google Patents

Abstract

PURPOSE:To prevent the image pattern of a character from protruding out of a frame indicating the character position by correcting the character position given by a format table in accordance with the measured result of a form length. CONSTITUTION:A light/darkness contrast between the back-ground on which a form 1 is placed and the surface of the form 1 is preliminarily set to a high value or proper positions in edge parts of the form 1 are preliminarily marked to detect coordinate positions of form edges with a scanner 4. Both ends in the X-axis direction and both ends in Y-axis direction are detected and are stored in a form end memories 71 and 72. Both ends in the X-axis direction and both ends in the Y-axis direction are preliminarily stored in the format table in a recognizing device 8, and the form length and the character position are obtained by a form length measuring instrument 13 and a correcting device 14. The recognizing device 8 recognizes characters in accordance with the obtained character position.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention provides an optical character reading device that reads characters written in predetermined positions on a form and determines the character code representing the character. It is related to.

[Conventional technology]

Within the data processing device, all characters are processed in the form of character codes. However, humans recognize characters from the image of the character, and the data processing device is equipped with a character generator, and when the character code of the character is input to the character generator C; It is configured to output a character image (referred to as a character image pattern) and display this character image pattern on a display device or print it out with a printer.

On the contrary, a device that generates a character image pattern by photoelectric conversion from characters printed or handwritten on a form and determines the character code of the character from this image pattern is called an optical character reader. and
An optical character reading device is used for the purpose of reading characters expressed in a form that humans can recognize, converting them into a character code form, and inputting the converted character code into a data processing device.

FIG. 3 is a block diagram illustrating a conventional device. In the figure, (1) is a form, letters A, B, and C are written at predetermined positions on the form (+1), (2) is a light source, (3)
is a sensor, and a scanner (
4). (5) is a scanner control device, (6) is a control device, 17) is an image memory, and (8) is a recognition device.

The sensor (3) has a large number of photoelectric conversion elements arranged in the horizontal direction (hereinafter referred to as the X-axis direction), and converts the intensity of the reflected light from the form (1) illuminated by the light source (2) into an electrical signal. Convert.

That is, optical information corresponding to one raster in the X-axis direction of form (1) is converted into an electrical signal and sent to the scanner control device! 11
51. The intensity of the reflected light from the white area on the form (1) is strong, and the intensity of the reflected light from each point on the character line is weak, and the intensity of the output electrical signal is determined according to this intensity. The electrical signal from the scanner (4) is converted into a binary signal and stored in the image memory (7) via the scanner control device (5) and the control device (61).

In addition, the scanner control device (5) also controls the
When you have finished inputting information for one raster in the axial direction, move the scanner (4) in the Y-axis direction perpendicular to the X-axis to input the next raster, and in this way, form the form (1). scans the entire surface of the 1m-order image memory (7
). The address position of each bit in the image memory (7) corresponds to the coordinate position on the form file of the pixel corresponding to the bit.

The recognition device (8) reads out the image pattern of each character from the image memory (7) and determines the character code corresponding to the image pattern by a method conventionally known as pattern 1dil&.

FIG. 4 is an explanatory diagram showing a comparison between the contents of the format table stored in advance in the catfish identification device fl+81 and the contents of the image memory (7) in order to make pattern recognition processing as easy as possible. , (9) is the coordinate position indicating the edge of the form paper in the format table and the form (
1) The coordinate position of the frame (10) indicating the position where the character should be written is tentatively displayed using a drawing, and (11) corresponds to (9) of the format table in the image memory (7). Address position if, (12) is stored in the image memory (7) in the frame (10) of the format table.
), and A, B, and C are the frames corresponding to the image memory (
7) is the address position where the image pattern of each character is actually stored.

Image memory (7) Rather than recognizing each character separately from the image pattern in the entire area, the character is recognized from the image pattern in each frame (10) in which each character is written. &i! Inside the 1m device (8) there is a frame (1) shown in Figure 4.
The coordinate position of 0) is stored as a format table, and character recognition is performed by reading image patterns only within the address range of the frame (12) corresponding to the frame (10).

[Problem that the invention seeks to solve]

However, the actual coordinate position of the character in the image memory (7) may not be the same as the one given in advance as representing the character position because the form expands or contracts due to changes in humidity, etc., or there is misalignment with the scanner (4). There was a problem in that it protruded from the coordinate position of the frame (12) and reached the relative position shown in FIG. 4, which hindered character recognition.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide an optical character reading device in which a character image pattern does not protrude from the coordinate position of a frame.

[Means for solving problems]

In this invention, a paper length measuring device is provided to measure the paper length of form paper.The ratio between the actually measured paper length and the paper length given by the format table is determined, and this ratio is used to determine the address position of the frame. was corrected so that the character image pattern did not protrude from the frame.

[Effect]

Since the character image pattern does not protrude from the frame, the character recognition rate is improved.

〔Example〕

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing one implementation of the present invention, and in FIG. 1, the same reference numerals as in FIG. 3 indicate the same or corresponding parts,
(13) shows the function as a paper length measuring device provided in the control device ft+61, (14) shows the function as a correction device provided in the recognition device (8), (71), (72)
) are paper edge memories. Figure 2 also shows the format table and image memory (
7), the same reference numerals as in FIG. 4 indicate the same or corresponding parts, and (12a) is the correction device (14).
) shows the coordinate position of the frame corrected by

Make sure that there is a good contrast between the background on which the form (1) is placed and the surface of the form (1), or
), the coordinate position on the paper can be easily detected by the scanner (4). Both ends in the X-axis direction and both ends in the Y-axis direction are detected and stored in paper end notes IJ (71) and (72).

Both ends of the X-axis direction and both ends of the Y-axis direction are recognition devices 1tist.
Since it is stored in the format table in the paper length measuring device [(13) and the correction device 1ft (14),
Constant ar b + expressed by the correction formula of ye-+21
The value of X6 t 76 can be determined. However, xo and yo in the formula fil, +21 represent the coordinate position on the format table, and X r )' represents the coordinate position on the image memory. The coordinate position of the frame (lO) is 2Fil, +2
1 to convert to the coordinate position of the frame (12a) on the image memory. The position of the frame (12a) exactly matches the position of the character image pattern ABC.

Therefore, by recognizing the image pattern of each character according to the coordinate position of the frame (12a), the recognition process becomes easier and the character reading rate can be improved.

〔Effect of the invention〕

As described above, according to the present invention, since the character position given by the format table is corrected according to the paper length measurement result, the character image pattern does not protrude from the frame indicating the character position, and the optical system has a high reading rate. A character reading device can be obtained.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is an explanatory diagram comparing the format table in the recognition device and the contents of the image memory in the device shown in FIG. 1, and FIG.
The figure is a block diagram showing a conventional device, and FIG. 4 is an explanatory diagram comparing the format table in the recognition device and the contents of the image memory in the device of FIG. (1) is a form, (4) is a scanner, (5) is a scanner system # device, (6) is a control device including a paper length measuring device It (13), (7) is an image memory, and (8) is a correction J&t
(14) A recognition device comprising: Note that the same reference numerals in each figure indicate the same or corresponding parts.

Claims (1)

[Claims] A scanner that scans a form on which characters are written in predetermined positions and converts reflected light from the form into an electrical signal, and information converted into an electrical signal by the scanner. a control device that converts the image pattern into an image pattern including an image pattern of each character on the form and stores it in an image memory; and a control device that reads the image pattern of the character stored in the image memory and determines a character code representing the character. an optical character reading device comprising: a means for reading the position of the edge of the paper of the form by the scanner; a paper length measuring device for calculating the paper length from the position of the edge of the paper; The calculated actual paper length and
A correction device that calculates a ratio to a predetermined paper length and uses this ratio to convert a predetermined character position corresponding to each character to an actual position of each character on a form; An optical character reading device comprising: means for recognizing each character in the recognition device according to the actual position of each character on the written form.