JPH01155490A - Optical character reader - Google Patents

Abstract

PURPOSE:To shorten a read time by deciding an identified result when the same identified result reaches more than half of the total number of screen memories. CONSTITUTION:Picture information stored in a picture information storage means 8a is transmitted to a position detection means D, and positions where a character, a symbol and the like exist are detected. Detected position information are stored in a detection position corresponding means E. Picture information which the detection means D is transmitted to a sequential identification processing means B, and the character, the symbol and the like are identified. When all picture signals stored in the screen memory 8a are identified, a similar processing is executed on picture information stored in a preceding picture memory 8b. In such a case, the identified result is decided when the same results among the identified results transmitted from the processing means B to a majority decision processing means C is detected to exist more than half the total number of the picture storage means.

Description

[Detailed Description of the Invention] [Field of Industrial Application] This invention stores image information including characters, symbols, etc. by repeatedly scanning an image sensor, and identifies characters, symbols, etc. based on the image information. The present invention relates to an optical character reading device, etc., and particularly to a hand-held optical character reading device that reads characters by holding a scanner and applying the scanner to a line on which characters are written, such as on a price tag.

[Prior art and its problems]

At supermarkets, department stores, etc., POS (Point O) collects sales information for each item and manages inventory.
f5ales) systems are widespread, and hand-held optical character reading devices are frequently used. FIG. 5 shows an example of the configuration of a conventional hand-held optical character reading device.

Hereinafter, its configuration will be explained based on FIG. 5. scanner 1
is held in hand 2 and applied to paper 3, such as a price tag used in a POS system, for example.

The scanner 1 includes photoelectric conversion elements arranged in a planar manner, and includes an illumination light source 4, a lens 5, an image sensor 6, and a drive/binarization circuit 7. The image sensor 6 has a field of view of at least one device of the characters written on the paper 3. The drive/binarization circuit 7 converts the analog signal, which is the output signal of the image sensor 6, into binary signals corresponding to the character area and the background area, respectively. The scanner 1 is operated, for example, by a push button switch. The binarized signal converted by the drive/binarization circuit 7 is stored in the screen memory ga.
, 8b, 8c. Screen memory 8a, 3b, 8c
is connected to the − digit memory 10 via the − digit extraction unit 9 . - The digit extraction section 9 is composed of an X coordinate counter 9a and a single digit extraction circuit 9b. - The digit memory 10 is connected to the single character memory 12 via the - character cutting section 11, and the - character memory 12 is connected to the character identification circuit 13.
It is connected to the.

FIG. 6 shows the configuration of the - character cutting section 11. The horizontal OR circuit 11a inputs the characters in the negative digit memory 10.
Find the Y coordinate value of the symbol, etc., and -character cutting circuit 11
Y for retrieving a single character image from the single digit memory 10 in b
Give coordinate values.

- The character cutting circuit 11b is connected to the Y from the horizontal OR circuit 11a.
The coordinate value is received, and a single character image is taken out from the minus digit memory 10 and stored in the single character memory 12.

The identification result buffers 14a, 14b, and 14c are connected to the X-coordinate counter 9a and the character identification circuit 13, and are configured so that predetermined data can be input thereto.

Identification result buffer 14a.

14b and 14c are a digit alignment processing section 15 and a majority decision processing section 1.
6, and the identification result is confirmed. Screen memory 8a
, 8b, and 8c store binarized data corresponding to the entire field of view of the image sensor 6 (for example, pXQ pixels when the X coordinate is - p1 and the Y coordinate - q). Note that since the character recognition circuit 13 described later recognizes one character at a time, it is necessary to retrieve data for one character from the screen memories 8a, 8b, and 8c. Therefore, the minus digit extraction section 9b is used as the single character extraction section 11 from the screen memories 8a, '8b, and gc.
The character extraction unit 11 extracts data corresponding to m×q pixels that can be processed by the single-digit memory 10 and stores it in the single-digit memory 10, and the character extraction unit 11 extracts data corresponding to m×q pixels that can be processed by the character identification circuit 13 from the single-digit memory 10.
Data corresponding to a pixel (data for one character) is extracted and stored in the one-character memory 12.

Hereinafter, a method for retrieving one character's worth of data from the screen memory 8a will be explained based on FIG. FIG. 7(a) shows data 3o read from the paper 3 by the scanner 1 and stored in the screen memory 8a. In the same figure (a), the - digit cutting part 9 is first cut out from Xl-1 to X2-m1.
The data from Yl-1 to Y2-q is taken out from the screen memory 8a and transferred to the negative digit memory 10 ((bl) in the same figure).
'). - The character cutting unit 11 judges the contents of the one-digit memory 10 and transfers n (n<q) portions of the range including the character image to the one-character memory 12 ((e,) in the figure). -Character memory 1
If 2 contains an identifiable character, character identification circuit 1
3, the character is recognized. Note that how to determine the range including the character image will be described later based on FIG.

Next, data from X, -2 to X 2- m +1, Y, , -1 to Y2-q is taken out from the screen memory 8a and transferred to the negative digit memory 10 (FIG. 7 (b2)). Then, the image in the range including the character image is transferred to the single character memory 12 ((C2) in the figure). Thereafter, in the same manner, the screen memory 8a
The image containing the character image is transferred to the one-digit memory 10 by sequentially shifting the position to be taken out from the character image, and the image including the character image is transferred to the one-character memory 12 and processed by the character recognition circuit 13 to perform recognition of one device. Note that the minus digit extraction section 9 consists of an X coordinate counter 9a and a single digit extraction circuit 9b.

The X-coordinate counter 9a indicates the coordinate when an image is taken out from the screen memory 8a to the one-digit memory, and its value increases successively as the - finger images are taken out. - The digit extraction circuit 9b has a function of extracting an image from the screen memory 8a to the one-digit memory 10 according to the extraction position indicated by the X-coordinate counter 9a.

Next, how to determine the range including the character image will be explained based on FIG. First, a row in the X direction is focused on, and if there is a black pixel in that row, it is set to 1, and if there is no black pixel, it is set to 0. For example, if the black output of the sensor is expressed as 1 and the white output as O, the result of the calculation will be the logical sum of each pixel in one row (hereinafter, such calculation will be referred to as "horizontal OR processing").
. Therefore, in the parts where there are letters! As shown in FIG. 8(b), the result of the horizontal OR processing is black for that range.

For example, if the horizontal OR processing becomes black from Y-13, the range to be transferred from the -digit memory 10 to the one-character memory 12 is n pixels from Ylll, including the white above the character.

Through the above processing, characters, symbols, etc. included in the field of view of the image sensor 6 can be read. The image sensor 6 is scanned and the image of the paper 3 is stored in the screen memory 8a,
8b.

The process of recognizing each character, symbol, etc. in the screen memory is sequentially stored in the 8C, and the recognition result of the character, symbol, etc. and the position of the character within the visual field (when the character, symbol, etc. is recognized) is The X coordinate when the screen is cut out into the single digit memory 10 by the single digit cutting unit 9 is the identification result buffer 14a, 14b,
Stored in 14c.

In addition, in the optical character reading device shown in FIG. 5, the identification rate is improved by taking a majority vote from the results of repeated identification of one character, and the identification result buffer y14
a + 14 b T 14 c s The digit alignment processing section 15 and the majority decision processing section 16 are means for taking a majority decision on the character identification results when the screen is captured and identified three times. First, the X-coordinate values of the characters stored in the identification result buffers 14a, 14b, and 14c and the identification results are sent to the digit alignment processing section 15, and the digit alignment processing section 15 performs the following operations based on the X-coordinate values of the characters: Match character identification results that can be determined to be the same digit. The majority decision processing unit 16 takes a majority decision between the matched character identification results and obtains the final identification result for that digit.

An example of majority decision will be explained below based on FIG. 9.

3(a) shows the lines written on paper 3.

FIGS. 9(b), 9(b), and 9(b3) show the identification results of the first, second, and third identifications, respectively. In the same figure (bl), the letter “1” is missing, in the same figure (b2), the letter “2” is missing, and in the same figure (b3), the letter “2” is missing.
), the character "3" is missing, but by matching the identification results for each digit based on the X coordinate within the field of view of the character and then taking a majority vote, the correct answer can be found as shown in the same figure (b4). It has been obtained.

Note that if you do not use the X coordinate of the characters and simply match them from the beginning of the identified characters and take a majority vote, the matching may fail and the correct answer may not be obtained, as shown in (C) in the same figure.
A digit alignment processing section 15 is provided. Therefore, even if there is a digit with a missing character, the 9th
As shown in Figure (b4), majority decision processing can be performed based only on the identification results for the same digit.

However, with conventional optical character reading devices, all the data in multiple screen memories is identified from the beginning, so when identifying multiple screens, the time required to recognize one screen is times the number of screens. . Therefore, there is a drawback that the reading time becomes long.

Therefore, this invention aims to determine the final identification result by taking a majority vote of the identification results obtained on each screen.
Identification results necessary for a majority decision to be determined by only the identification results of some images for symbols, etc. (for example, if the number of images is 3)
If 2) is obtained, the characters and
Focusing on the fact that identification processing of symbols, etc. is unnecessary, the reading time is shortened by finalizing the identification results when the same identification results reach a majority of the total number of screen memories.

[Means for solving problems]

In order to solve the above problems, the present invention provides an optical character reading device that stores image information including characters, symbols, etc. by repeatedly scanning an image sensor, and identifies characters, symbols, etc. based on the image information. , an image information storage means for storing a plurality of pieces of image information as binary information; a position detection means for detecting a position where a character, symbol, etc. is present based on one piece of binary information stored by the image information storage means;
a detection position correspondence means for correlating the position where a character, symbol, etc. detected by the position detection means exists with other binary image information; and a position detection means for detecting the position from the binary image information stored in the image information storage means. an identification processing means for identifying and processing image information such as characters and symbols cut out while being handled by the detection position handling means;
majority determination means for detecting that one identification result for the symbol etc. has reached a majority of the total number of binarized image information stored in the image information storage means; and majority voting processing means for specifying a plurality of identification results including the identification result obtained by a majority vote into one identification result, and when one identification result reaches a majority, majority voting processing is performed and the identification result is specified. It is characterized by

[Effect]

Since the present invention is configured as described above, by the interaction between the majority determining means and the majority processing means, when one identification result reaches a majority, the majority processing is performed and the identification result is specified.

〔Example〕

DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an optical character reading device according to the present invention will be described below with reference to the accompanying drawings. In the description, the same reference numerals are used for the same elements having the same functions (actions) as the constituent elements of the invention, and redundant description will be omitted.

FIG. 1 shows an example of the basic configuration of an optical character, etc. reading device according to the present invention. Scanner 1 to screen memory 8a
, 8b, and 8c are the same as those in the prior art, so their explanation will be omitted. This basic configuration example basically includes image information storage means A1 identification processing means B1 majority decision processing means C1 position detection means D1 detected position correspondence means E1 and majority determination means F.

The image information storage means A is connected to the position detection means and can transmit stored image information to the position detection means. The position detection means is the detection position corresponding means E.
and identification processing means B. Therefore, the position detection means can send the detected position information to the detected position correspondence means E, and can also receive the position information sent from the detected position correspondence means E. Identification processing means B
identifies characters, symbols, etc. based on image information including characters, symbols, etc. obtained by the position detection means.

Identification processing means B is connected to majority decision processing means C, and majority decision processing means C specifies correct characters, symbols, etc. by majority decision based on a plurality of identified identification results. The majority decision means F is connected to the majority decision processing means C and the detection position correspondence means E, and decides whether or not the identification results sent to the majority decision processing means C have reached a majority, and if the determination is more than a majority. In some cases, a processing signal is sent to majority decision processing means C to perform majority decision processing. On the other hand, if the determination is less than a majority, a signal is sent to the detection position handling means E to continue the identification process. Here, the majority depends on the total number of image information storage means A. If the total number is N, the majority is (N+1)/2 if N is an odd number, and N/2+1 if N is an even number. The detected position corresponding means E stores the following image information storage means 8.
Characters, symbols, etc. are identified by associating the image information stored in b with the detected position. FIG. 2 shows an example of a concrete connection method of the majority decision processing means C and the majority judgment means F. In this case, since the total number of screen memories is three, the majority is two, and if there are two identical results, the majority result can be determined. Therefore, the identification result buffers 14a, 14b and the majority determination means F are connected, and the identification results of the first screen and the second screen are sent to the majority determination means F.

Next, the operation of this embodiment will be explained. The image information stored in the image information storage means 8a is sent to the position detection means, and the positions where characters, symbols, etc. are present are detected in the Y direction and the X direction. As described above, the detection in the Y direction detects the position where a character, symbol, etc. is present by calculating the horizontal OR process. Detection in the X direction detects the position where a character, symbol, etc. is present while sequentially shifting the position by a certain range. What is important here is that the position information detected by the position detection means, that is, the position information where characters, symbols, etc. are present, is stored by the detected position correspondence means E. The image information detected by the position detection means is sequentially sent to the identification processing means B, where characters, symbols, etc. are identified. When all the image information stored in the image information storage means 8a has been identified, the same processing as described above is performed on the next image information stored in the image information storage means 8b. In this case, the difference from the above processing is that position information is sent from the detected position correspondence means E to the position detection means, and detection in the image information is performed based on the position information stored in the detected position correspondence means E. Among the identification results sent to the majority decision processing means C, the identification result is determined when it is detected that the same result exists in more than half of the total number of image information storage means. Therefore, in the image information storage means 8a, xl-5, Y
If information such as characters and symbols is identified in l-11,
The identification process in the image information storage means 8b and 8c is
-5, Yl-11, for example, when the image information storage means 8a and 8b are identified, if there are two identification results of "0", that is, more than a majority, majority voting processing is performed. In this case, the majority vote result is "0".

Next, a first embodiment of the optical character reading device according to the present invention will be described with reference to FIG. First, the correspondence with the basic configuration example shown in FIG. 1 will be explained.

In this case, the screen memory 8a serves as the image information storage means A.
, 8 b r 8 Cs The character recognition circuit 13 serves as the identification processing means B, and the identification result buffer y serves as the majority decision processing means C.
14a, 14b, 14c, digit alignment processing unit 15 and majority decision processing unit 16, single digit extraction circuit 9 as position detection means
b, -digit memory-0, -character cutting unit 11 and -character memory-2, X coordinate counter 9a as detection position corresponding means E, coordinate value switching circuit 17, coincidence judgment circuit 20 as majority judgment means F; There is.

The difference from conventional optical character reading devices is that it includes a coordinate value switching circuit 17 and a match determination circuit 20. The coordinate value switching circuit 17 is connected between the X coordinate counter 9a and the one-digit extraction circuit 9b, and the coincidence determination circuit 20 is connected to the digit alignment processing section 15 and the coordinate value switching circuit 17. Here, coordinate value switching circuit 1
Reference numeral 7 is for switching the value of the X coordinate given to the single digit extraction circuit 9b, and the matching judgment circuit 20 is used to determine whether one identification result and the other identification result are the same when the character identification circuit 13 identifies the presence of a character. This is to determine whether or not the Here, "when the presence of a character is identified" means when the character identification circuit 13 identifies a character or outputs an unidentifiable message. Furthermore, "unidentifiable" means that although characters exist, it is not possible to identify what characters they are. In this case, if the identification can be made correctly on the second and third screens, the correct answer can be obtained through majority voting.

Next, the operation of the first embodiment will be explained. The identification results on the first screen, including cases where identification is not possible, are stored in the identification result buffer 14.
It is stored in a. In the identification result of the first screen, the position where the single-digit clipping circuit 9b cuts out the single-digit image is the coordinate value output by the X-coordinate counter 9a. Since the X coordinate counter 9a outputs a value in which the X coordinate is incremented by 1, in the case of the first screen, one digit is extracted from the entire screen regardless of the presence or absence of characters. When identifying the second screen of the screen memory 8b,
This is done in the same way as the first screen.

The identification results of the character identification circuit 13 on the second screen are stored in the identification result buffer 14b, and when the identification processing on the first and second screens is completed, the identification results of each character are
A process of determining whether or not identification of the third screen is necessary is performed. First, the digit alignment processing unit 15 displays the first screen and the second screen.
The digits of the identification results on the screen are aligned, the identification results after digit alignment are set as Ra and Rb, and the X coordinate of the digits is set as Xi.

The majority decision result can be determined from the above two identification results as follows.
This is the case of Ra-Rb. Therefore, the match determination circuit 2o determines whether Ra and Rb match. If the two match, the majority decision processing unit 16 obtains the majority decision result using only these two identification results (if Ra-Rb,
The majority vote result is Ra). If they do not match, the third screen stored in the screen memory 8c is identified using the coordinate values of Xi obtained from the digit alignment result.

The X coordinate value used at this time is determined by the coordinate value switching circuit 17.
The X coordinate value is switched from the X coordinate counter 9a to the X coordinate value output by the digit alignment processing section 15.

When the identification result Rc of the third screen at the coordinates of Xi is obtained from the character identification circuit 13, it is set in the identification result buffer 14c, and the majority decision processing unit 16 receives Ra.

A majority decision process is performed based on Rb and Rc. As a result, characters can be identified on the third screen only by the X coordinate where the characters exist, and as with the second screen, the screen can be recognized in a much shorter time than on the first screen. can be identified.

In this embodiment, horizontal OR processing for extracting - characters is performed on all characters, symbols, etc. existing in the - line, but by providing a Y coordinate counter, a coordinate value switching circuit, and a coordinate buffer, Y where characters/symbols, etc. exist
Coordinate values can also be stored.

In this case, a coordinate value switching circuit is connected to the - character cutting section 11, and a Y coordinate counter is connected to this coordinate value switching circuit. Furthermore, a coordinate buffer is connected between the Y-coordinate counter and the coordinate value switching circuit, so that the coordinate value of the Y-coordinate counter can be stored in the coordinate buffer. Since the functions and operations of each are as described above, explanations thereof will be omitted. With this configuration, in the digit containing the second extracted character/symbol, etc., the position where the character/symbol, etc. was detected in the first digit (Y, -11 (see Figure 7 (c)))
Since it is possible to cut out characters, the time required for identifying characters, symbols, etc. can be shortened.

Next, a second embodiment of the optical character reading device according to the present invention will be described based on FIG. 4. First, the correspondence with the basic configuration example shown in FIG. 1 will be explained.

In this case, the screen memory 8a serves as the image information storage means A.
r 8 b * 8 Cs character identification circuit 13 as identification processing means B, identification result buffers 14a, 14b, 14c and majority decision processing unit 16 as majority decision processing means C, single digit extraction circuit 9b as position detection means, -digit memory 1
0, -character cutting section 11, -character memory 12, X coordinate counter 9a and screen switching control circuit 19 as detection position correspondence means E, and coincidence judgment circuit 20 as majority judgment means F.

The difference from conventional optical character reading devices is that it has a screen switching control circuit 19 and a match determination circuit 2o, but does not have a digit alignment processing section 15 (see FIG. 5). The screen switching control circuit 19 has a function of switching the screen memory, and the screen memory 8a.

8b, 8cSX coordinate counter 9a and character identification circuit 13
It is connected to the. Therefore, the character recognition circuit 13 is configured to receive a screen switching signal and send a count stop signal to the X coordinate counter 9a. The match determination circuit 2° includes identification result buffers 14a, 1
4b, majority decision processing unit 16 and screen switching control circuit 1
9 is connected. Therefore, the identification result buffer 14
a.

The identification result is received from 14b, and if the two results match, a processing signal is sent to majority decision processing section 16 for majority decision processing. On the other hand, if the two results do not match, a screen switching signal is sent to the screen switching control circuit 19 in order to continue the identification process.

Next, the operation of this embodiment will be explained. Scanner 1 in hand 2
The image of characters, symbols, etc. illuminated by the illumination light source 4 is formed on the image sensor 6 by the lens 5. The image sensor 6 is scanned by a drive/binarization circuit 7, and the binarized images of characters, symbols, etc. are stored in screen memories 8a, 8b.
, 8c. - The digit extraction section 9 is the screen memory 8
Sequentially cut out the -digit image from a and send it to the single-digit memory 10,
- The character cutting unit 11 cuts out a single character image from the single digit memory 10 and sends it to the single character memory 12. -Character memory 12
The single character image sent to is identified by the character identification circuit 13. When the character identification circuit 13 identifies the presence of characters, symbols, etc., a signal is sent to the screen switching control circuit 19.

The screen identification circuit 19 sends a count stop signal to the X-coordinate counter 9a, and at the same time, the negative digit clipping circuit 9b switches the target from which the one-digit image is clipped to the screen memory 8b.

Identification processing of characters, symbols, etc. in the screen memory 8b is performed at the position (Xl-5) where the characters, symbols, etc. were present on the first screen by the action of the detection position correspondence means E. The identification result Ra from the first screen and the identification result Rb from the second screen are stored in identification result buffers 14a and 14b, respectively. Thereafter, the match determination circuit 20 determines whether the two identification results match. If they match, the identification process for the third screen is not performed, and the majority decision processing section 16 outputs Ra (-Rb) as the majority decision result. If they do not match, a screen switching signal is output, the screen switching control circuit 19 holds the value of the X coordinate counter 9a, and the minus digit cutting circuit 9b switches the cutting target to the screen memory 8c. Through this process, the screen memory 8a is stored at a position corresponding to a position where a character, symbol, etc. exists in the screen memory 8a.
c can be identified. The identification result Re obtained from the third screen from the character identification circuit 13 is stored in the identification result buffer 14c, and Ra, Rb, and Re are sent to the majority decision processing section 16.
The majority result of the three identification results is output as the final identification result. When the majority result is determined for that digit through the above processing, the screen switching control circuit 19 releases the stopped state of the X-coordinate counter 9a, and returns the cutout of the - digit image to the first screen of the screen memory 8a. Thereafter, the identification process on the first screen continues until the presence of the next character, symbol, etc. is identified by the character identification circuit 13.

Note that when capturing three screens, if the relative positions of scanner 1 and paper 3 shift and there is a possibility that the X coordinate value of the same character image on each screen may shift slightly,
Separate the characters, including the vicinity of the character position determined by screen identification. For example, if the predetermined character coordinates on the first screen are Xl,
On the second screen, X -1, X SX for the character
+1 3 or l 1 l Identify in place. In this case, the processing time is three times longer than when only Xl is identified on the second screen, but the processing time can be shortened compared to when identifying at all locations on the screen as in the conventional technology. can.

Since this embodiment is configured as described above, the characters and
The positions of symbols, etc. are determined only on the first screen, and if the identification results are the same, the identification results are determined on the first screen and the second screen, so the processing time can be shortened.

〔Effect of the invention〕

Since the present invention is configured as described above, the time required for identifying image information can be shortened with a simple circuit configuration.

In particular, since the X-coordinate counter used in conventional optical character reading devices can be used, the number of additional parts can be reduced.

[Brief explanation of the drawing]

FIG. 1 is a diagram illustrating a basic configuration example of an optical character reading device according to the present invention, and FIG. 2 is a diagram illustrating a connection example between majority processing means and majority determination means in the basic configuration example of FIG. 1. , FIG. 3 is a diagram showing a first embodiment of the basic configuration example in FIG. 1, FIG. 4 is a diagram showing a second embodiment of the basic configuration example in FIG. 1, and FIG. 5 is a diagram showing a second embodiment of the basic configuration example in FIG. FIG. 6 is a diagram showing the configuration of a conventional optical character reading device; FIG. 6 is a diagram showing the configuration of a character extraction section; FIG. 7 is a process diagram for explaining the process up to character extraction; FIG. is a diagram for explaining a - character extraction method, and FIG. 9 is a diagram showing an example of majority decision. A... Image information storage means B... Identification processing means C.
...Majority processing means D...Position detection means E.
...Detected position corresponding means F...Majority determination means 1...
・Scanner 2...Hand 3...Paper
4...Illumination light source 5...Lens 6
...Image sensor 7...Drive ψ binarization circuit 8...Screen memory 9...-Digit extraction unit 1
0...-digit memory 11...-character cutting section 12...-character memory 13...character identification circuit 1
4... Identification result buffer 15... Digit alignment processing section 16... Majority decision processing section 1
7...Coordinate value switching circuit 18...Coordinate buffer 19...Screen switching control circuit 20...Coincidence judgment circuit Patent applicant Sumitomo Electric Industries, Ltd. Patent attorney Representative Yoshiki Hase Yoshikima
Mountain 1) Line - Procedural amendment January 14, 1988 1 Indication of the case Patent Application No. 315013 of 1988 2 Name of the invention Optical character reading device 3 Person making the amendment Relationship to the case Patent applicant (213 ) 4th Director, Sumitomo Electric Industries, Ltd. (Postal code 101) 4th floor drawing of Foresight Building, 3-5-2 Iwamoto-cho, Chiyoda-ku, Tokyo

Claims (1)

[Scope of Claims] 1. An optical character reading device that stores image information including characters, symbols, etc. by repeatedly scanning an image sensor, and identifies the characters, symbols, etc. based on the image information, comprising: an image information storage means for storing a plurality of pieces of information as binary information; a position detection means for detecting a position where a character, symbol, etc. is present based on one piece of binary information stored by the image information storage means; detection position matching means for associating the position of the character, symbol, etc. detected by the position detection means with other binary image information; an identification processing means for identifying and processing image information such as characters, symbols, etc. cut out by the position detection means while being matched by the detected position correspondence means; and an identification result for the characters, symbols, etc. obtained by the identification processing means. , majority determination means for detecting that the total number of binarized image information stored in the image information storage means has reached a majority; and based on the detection information of the majority determination means, an identification result that has reached at least the majority; and a majority decision processing means for specifying a plurality of identification results including a plurality of identification results as one identification result by majority vote, and when the one identification result reaches the majority, the majority decision processing is performed and the identification result is specified. Characteristic optical character reading device. 2. The detected position corresponding means is a storage means for storing the position where a character, symbol, etc. is detected when the position detecting means identifies the first binary image information, and the position detecting means When switched to the target binary information,
The optical character reading device according to claim 1, which cuts out characters, symbols, etc. based on the stored position information. 3. The majority decision processing means has a digit alignment processing section that performs a digit alignment process on the identification result based on the position information detected by the position detection means, and performs a majority decision process on the identification result that has undergone the digit alignment process. An optical character reading device according to claim 1.