JPS58101370A - Character recognition device - Google Patents

Abstract

PURPOSE:To attain reading state by setting reading paper one by one to a prescribed location, through the recognition of characters printed on a reading objective plane from an image signal a scanning section. CONSTITUTION:When a reading receptionable state signal is transmitted from an external device to a control section 21 with a signal 281 through an interface 28 and a signal 212, the control section 21 moves a drive motor 19 to a reading start position with a signal 211 to display the possibility for reading to an operating button 10 with a signal 210. After confirming this display, a paper 12 is set to a reading window 11 and th start of reading is instructed with the operating button 10. When the control section 21 detects the start of reading via the signal 210, the reading for one line's share is started by driving the drive motor 19. As the result of reading, a character code, a value of reliability for code discrimination, as required is outputted to an external device with the interface 28 and the signal 281.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides an optical character reading device (Qptical Ch.
a-yactey shader, hereinafter abbreviated as OC
It is related to ``Band''.

Conventionally, a postal code reader or gap is a device that converts characters printed or written on paper into electrical signals using a photoelectric conversion system, and uses various recognition methods to determine the character code corresponding to the character. It has been developed and commercialized as a compact OCR that uses an OCa or a hand scanner as an input means. These two types of devices are suitable for inputting relatively large amounts of data by reading the paper to be read using a conveyance system, and therefore belong to high-end devices and are inexpensive. It was hard to say. In addition, the small WOCR41 input system using a /% scanner as an input means is manually operated, so it is not suitable for long lines in the line direction, and there are often restrictions on the reading target and reading performance.

On the other hand, with the spread of computers, requests for inputting text information tend to be diversified and distributed, as typified by personal computers, etc., and there is a need for compact, paper storage, and highly reliable OC assistants. It has reached the point where

The object of the present invention is to develop a compact and inexpensive O
Our mission is to provide CR+.

According to the present invention, an OC card having reading performance comparable to that of a conventional high-end machine can be obtained by simply setting the reading paper one sheet at a predetermined position. This will be explained in detail.

FIG. 1 is a diagram showing an overview of the device to be invented, and is distinguished into scanning &lS1 and -w&82 for easy understanding.These may of course be housed in the same housing.Scanning system 1 includes a reading window 11 through which a piece of paper 12 to be read is to be sectioned, and a light-shielding power/(-
13, further provided with an operation button 10 for turning on the power, starting trial sampling, etc.

Figure 2 is for explaining the inside of the scanning system. It is a schematic diagram seen from one side. The reading window 11 at the top of the figure, the piece of paper 12,
The light shielding cover 13 corresponds to that shown in FIG.

The vL handle 11 is a flat plate made of glass or acrylic resin, and is designed so that the characters read on the piece of paper can be observed from below. Therefore, the piece of paper must be turned over and set in the reading window. The receiver IJli 16, which has an optical lens 161 and a light receiving element 162 as its main elements, is fixed on a movable base 17f that is movable in the left-right direction in the figure by a screw shaft 18 rotated by a motor 19. In addition, there are 14 lighting lights on the movable platform.
and mirror 15 are fixed in a position defined by the optical design. The illumination lamp 14 is a light source elongated in the direction perpendicular to the drawing, and is e.g. a fluorescent lamp, which provides uniform illumination in the row direction of the page. 9 images are formed on the light-receiving element by the load'-coupling element).

In this case, the light-receiving element is perpendicular to the drawing and can receive light dimensionally in the paper strip direction, so in order to scan the entire paper surface, the movable base 17 can be moved one-dimensionally in the left-right direction. This operation is performed by the motor 19 screw shaft 18
It can be easily realized by The advantage of this configuration is that the paper surface 12, which is the surface to be weaved, is easily fixed by the M viewing window 11, and that the illumination unevenness caused by the illumination 17 is reduced when the movable table 17 moves in the fixed right direction for scanning. The reduction rate of the optical system is approximately 100 µm because the scanning resolution of the conventional OCR is 50 to 100 µm on paper, but the element pitch of the commercially available light receiving element is 13 to 14 µm. ''/3.57 or more, and a copying machine based on the reduction ratio l does not require any optical path length, and by using the mirror 15, the height of the device can be easily reduced and the device can be made smaller. Two-dimensional scanning of the paper surface can be obtained by simply moving the device one-dimensionally in a direction perpendicular to the light-receiving area of the device, and the design of the drive unit can be simplified because the thing fixed to the movable table is fixed. Since the conveyance system is omitted, restrictions on the object to be read are reduced, there is no drive mechanism or control circuit for the conveyance system, and there is no damage to the object to be read during conveyance.

The third factor is a block diagram illustrating the configuration of the recognition unit and its relationship with the scanning unit. Drive motor 1 configuring the scanning section
9 and the receiving section 16 which converts the row direction of the paper strip r2 into a one-dimensional linear light percentage, the M-Sakiku 2 has a control section 21. which performs overall control. A/D conversion s22 that converts the image signal 160 received by the image receiving unit 16 into an electrically quantized signal, a line buffer 23 that holds the image signal for -characters, and a cutting unit that cuts out the image signal for one character from the line image signal. 24.1 characters mm
Feature extraction unit 25 that extracts features according to the recognition method from the code.
, a dictionary 27 that stores discrimination criteria that serve as standards for character code discrimination. ◎ The configuration of these g iron parts 2 is as follows: It is the same as many of the basic configurations of the OC that have been developed in the past, and detailed description thereof will be omitted.

The overall operation is as follows: When the external device'##power) sends the control ready state signal to the control unit 21 via the signal 281η), the interface 1128 and the signal 212, the control unit 2
1 moves the drive motor 19 to the lt-take starting position by the signal 211, and also sends the signal 21 to the operation button 10.
0 displays the value. After confirming this display, the operator sets the piece of paper 12 in the reading window 11 as shown in FIG. 1 and instructs the start of reading using the operation button IO. When the control unit 21 detects the start of reading via the signal 210, it starts reading one line by driving the drive motor 19. The reading result is outputted by the interface 28 to the external device 1a via a signal 281, including a character code and, if necessary, a confidence value for code discrimination.

In this way, the explanation of the present invention has been completed.The explanation of the present invention has been completed.
However, using a timing belt in place of the screw shaft 18 and the drive motor 19, or using a wire link mechanism as used in the European Thai writer does not change the gist of the present invention.
! It's not something you do. Also, CCTxI as a light receiving element
) Instead, a light emitting element with an integrated illumination system is used using a half-body photoreceptor array, and rod lenses integrated in a straight line provide illumination 14, mirror 15, and optical lens 16.
Substituting T for 1 does not change the gist of 4: Kinuta.

[Brief explanation of the drawing]

Figure 1 is a diagram explaining the approximate cost of the device according to the present invention, and Figure # and Figure 2 are diagrams explaining the 8-seizure of the running capital l.
The third example is a block diagram illustrating the configuration of 11 Zheng 2 and its series with Hiki Fengfu 1. In the figure numbers, 1 is the running cotton, 21 knee sagittarius, 10 is the sink operation button, 11 is the poor mine yield, 121 is the soft piece, 13 is the light shielding cover, 14 is the 9-115 mirror, and 16 is the receiver. area, 1
7 is a movable platform, 18 is a screw shaft, 19 is a motor, 21 is WtlJ-Zheng, 22 is A/Uk Abe, 23 is Line Buff YT241's Kirita Shizheng, 25 is FuilL Hakutabe, 26 is Kanjoto , 27 is a silent oath. 28 indicates Inmetsu Ace Department, respectively.

Claims (1)

[Claims] By manually installing each target surface on a transparent flat plate with the surface to be read facing downward, a primary light provided at the bottom of the transparent flat plate and arranged so as to receive light in one direction of the target to be read; A driving mechanism that drives the original light-receiving element array and the illumination and optical coupling mechanism in a direction perpendicular to the above-mentioned scanning section that scans and receives an image of the entire surface to be sampled, and an image signal from the scanning section. A small desktop character recognition device characterized by comprising: a g-oribe for recognizing characters printed or written on a target surface.