JPS60181987A - Optical reader of character - Google Patents

Abstract

(57) [Abstract] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field of the Invention The present invention relates to an optical character reading device that optically reads alphabetic characters, numbers, symbols, etc. printed on forms, price tags, etc.

Technical Background and Problems Generally, in this type of device, a lamp is used to illuminate a medium on which characters or the like are printed, and it is determined whether or not the medium is a character based on the amount of reflected light. Therefore, even if it is obvious to the human eye that the text is not a text, it is often recognized as a text if there is a change in the amount of reflected light.

In particular, when the media is a price tag or the like, there is little space between the edge of the paper and the printed data, so it is easy to determine that shadows appearing at the edges of the paper are characters. Furthermore, in the case of paper in a plastic bag or coated paper, a different reflected light than usual is often generated, and it is easy to judge that this is also a character. In this case, even if this phenomenon appears in the normal data part, it will not be affected much because the change in the amount of reflected light in the normal data part is large.
Both ends of the data are often directly affected by this phenomenon and are output as characters. or,
Even when characters or symbols that are not intended for reading are printed near the characters to be read, noise tends to appear at both ends of the read data.

For this reason, one or more unnecessary erroneous character groups are likely to appear at both ends of a series of input data character strings.

There is a technique described in Japanese Patent Publication No. 58-36391 as a means for eliminating the influence of the erroneous character groups at both ends. In other words, a signal caused by the edge of paper that occurs at both ends of normal data is accepted as a signal, but it is an unidentifiable signal, and if the format checker recognizes this as an unidentifiable signal, it is determined that it is a series of data. Only the characters at both ends of the string are removed. Therefore, the influence of erroneous reading of both ends of a series of data character strings is prevented by means of removing unidentifiable signals.

However, in reality, when misreading occurs due to the edges of the paper, etc., rather than being recognized as unidentifiable characters, - (minus), ・(period), ・(
It is often read as a symbol such as a comma) or a space. Therefore, if it is read as such a specific character, it will not be added to normal data.
Even though normal data is being read correctly, an error occurs because the digits do not match.

OBJECTS OF THE INVENTION It is an object of the present invention to provide an optical character reading device that can improve the reading rate by removing noise without damaging the true data content.

Summary of the Invention The present invention provides special characters t that can be read as recognizable characters such as 1 (minus), .
- set in advance so that when this occurs at both ends of a series of data strings, they are removed,
This ensures that true data is not lost and no error signals are generated, thereby eliminating the effects of noise on both ends and improving the reading rate. be.

Embodiments of the Invention First, the scanner (1) is a handy type that scans a character string and optically inputs a plurality of characters sequentially from one direction. 3) and a lamp (4), and a COD (5) as a photoelectric conversion element is provided inside.

Furthermore, an LSI (6) and circuit components (7) for the COD (5) are built-in. Scanner like this (1)
are sequentially connected to a quantization circuit (8), a detection extraction circuit (9), a sampling circuit, a similarity calculation circuit α, and a list processing/editing circuit (2).

The optically read analog data as shown by arrow A is converted into digital binary data by the quantization circuit (8) as shown by arrow B.

Such binary data changes as shown in the figure within the detection frame α1. Assuming that the data for "2" is recognized, the signal in the detection frame αj changes sequentially from the left to the right in the diagram, and is extracted by the detection extraction circuit (9). The data is then cut out and normalized into data of 16 bits vertically and 12 bits horizontally.

Next, the recognized character data output from the list processing/editing circuit 4 is input to the I10 port α4. This I10
The port is connected to the CPU αυ, and this CPU (to) is connected to the ROMα0, RAMαη, counter α, and I10 port α that outputs to the host via the data bus (A) and address bus Qυ. has been done. Further, the counter (18 chip select (CS)) is connected to the I10 port α, ROM Qd, RAMaη, I1
Connected to port 0 (to).

As a medium, for example, there is a price tag (a), and the data A 123-456 is printed on this price tag (a). Specific characters that can be read as meaningful symbols due to the edge (a) of a price tag (a) or dirt, etc. are - (minus), ・(period), ・(comma),
Assume that it is U (space). These specific characters were selected because they were likely to occur when actually tested in various cases.

Next, the operation will be explained based on FIG.

First, after initialization, input data is accepted, its characters are recognized, and it is checked whether all characters have been input. When all characters are input, n indicating the number of digits is set to 1, and it is checked whether the first digit from the left end of all input characters is a specific character or not, and if it is a specific character, that character is deleted. Then, the process is repeated until a character other than a specific character is recognized as nt-2. For example, assume that the contents of all the characters read on the price tag (a) in FIG. 3 are -l L-IA123-456-.-p. Since the true data is A123-456, an error would occur if the corresponding processing was not performed. Therefore, as a result of performing the above-described processing, three digits 1-1 from the left side are removed.

However, when the character A is recognized in the processing on the left, it is not a specific character, so this time the number of digits from the right is n t
-1, and checks whether the first digit from the right side of all characters input is a specific character. If it is a specific character, remove that character, then add +1 to n and enter the second and third digits.
Check the digits in sequence. In the above example, from the right side t-
・The 1-digit character is removed.

When the left and right characters are removed in this way, the data becomes A 123-456 . After this data is subjected to various function checks, if there is an error due to other reasons, an error signal is output, and if normal, the data is output. Such noise deletion operation is performed by the list processing and editing circuit.

Effects of the Invention The present invention provides -(
If it is read as a specific character such as minus), ・(period), t(k-), -(space), etc., those specific characters may be one digit or consecutive multiple digits at both ends of the entire string. Since noise is removed on the condition that it exists, noise can be effectively removed and read without damaging the true data content, thereby improving the reading rate.

[Brief explanation of the drawing]

The drawings show an embodiment of the present invention; FIG. 1 is a block diagram, FIG. 2 is a block diagram of the circuit centering on the CPU, FIG. 3 is a plan view of a price tag, and FIG. 4 is a flowchart. . 5... CCD (photoelectric conversion element), 8... quantization circuit, 9... detection extraction circuit, 10... sampling circuit, 1
2...List processing editing circuit, 22...Price tag (media) application person Tokyo Electric Co., Ltd.

Claims (1)

[Claims] a photoelectric conversion element that optically reads characters or symbols printed on the media and converts them into electrical signals; a quantization circuit that converts the output of the photoelectric conversion element into a binary digital signal; A detection extraction circuit that extracts image data, a sampling circuit that eliminates the effects of noise, etc. in the image data, and a sampling circuit that checks the format of the entire input data and one or more specific symbols set in advance at both ends of the input data. An optical character reading device characterized by a list processing/editing circuit that removes and outputs specific symbols when they are continuous.