JPS6275885A - Character reader - Google Patents

Abstract

PURPOSE:To reduce the writing load of a character writer by writing simple and general-purpose characters to be converted in place of the same characters as those recorded on a form and recognizing these same characters. CONSTITUTION:A carrying/scanning part 11 carries a form to be read and a character pattern that is obtained by a photoelectric converter on the form is delivered to a preprocessing part 12. The part 12 detects characters one by one out of a group of character patterns delivered from the part 11 and performs noise processing and normalization processing. A character recognizing part 13 recognizes the character pattern for each character delivered from the part 12. A post-processing part 14 stores the results of recognition delivered from the part 13 by an amount equal to a single form and processes the result of recognition of the same recorded character designated by the character to be converted in the form of the result of recognition to the characters to be converted. A control part 15 stores previously various types of information needed for processing at the part 14 and performs the overall control of a character reader.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to a character reading device having a function of automatically outputting as a recognition result the same character as a recorded character designated by a character to be converted.

[Technical Background of the Invention and Problems Therewith] Conventionally, in an optical character reading device capable of reading handwritten kanji, for example, as shown in FIG. When characters are written, only each character is recognized. Here, when writing a large number of characters with the same content such as a name, address, etc. on the form 10, the characters may be completely the same. Even in this case, as shown in FIG. 5, it is necessary to enter the same characters each time. Therefore, if the same characters can be recognized in the same way as when the same characters are written by simply writing the characters or symbols, the workload of the person writing them can be reduced.

As a device that can solve this problem, there is an optical character reading device for alphanumeric and kana characters rlH that uses duplication/repeat symbols. Shikajinaga tl, this, 11
In the type device, there is only one heavy or repeated Nido R group,
The drawback is that it can only be applied to consecutive cases such as the next character in the first line. Furthermore, the decove and repeat symbol scheme is
The reading process for Kanji characters is quite appropriate.

An object of the present invention is to enable recognition processing to be performed on the same characters as those recorded on a form using a simple and versatile character entry method. The object of the present invention is to provide a character reading device that can reduce the input negative 1 [1] by a person writing characters.

[If invention! E Required] The present invention provides converted character information that instructs conversion into the same recorded character recorded on a form, and index field information that designates the recording field in which the same recorded character specified by this converted character information is recorded. and storage means for storing in advance field identification information for identifying recording fields. When the control means detects the recognition result of the character recognition means for the converted character recorded in a predetermined recording field of the form based on the memory contents of the storage means, the control means identifies the recording field of the converted character and records this recording. It is configured to output the same recorded character in the recorded field on the form designated by the index field information with the same recorded content as the field as the final recognition result of the converted character.

With this configuration, by writing a character to be converted, it is possible to obtain the same recognition result as when the same recorded character specified by the character to be converted is recognized.

[Embodiment of the invention] An embodiment of the invention will be described below with reference to the drawings. 1st
The figure is a block diagram showing the configuration of an optical character reading device according to an embodiment. In FIG. 1, a conveyance scanning unit 11 conveys a document to be read, and scans the document with a photoelectric converter to convert the obtained character pattern to a preprocessing unit 11.
Output to 2. The preprocessing unit 12 performs a check process, a noise process, and a normalization process for each character from the character pattern group output from the conveyance scanning unit 11. The character recognition unit 13 is
Recognition processing is performed for each character pattern output from the preprocessing section 12.

The post-processing unit 14 receives the recognition output from the character recognition unit 13.
This circuit stores the results for one form and processes the recognition result of the same recorded character designated by the character to be converted as the recognition result for the character to be converted. The control unit 15 is a circuit that controls the entire apparatus, and stores in advance various information necessary for processing by the post-processing unit 14.

In the apparatus N having such a configuration, the operation of this embodiment will be explained with particular emphasis on the post-processing section 14 and the control section 15. First, on the form 20 to be read, address fields 21a to 21c and name fields 22a to 22c for each recorded content are printed in dropout colors, as shown in FIG. 2, for example. Each field 21a to 21c
, 22a to 22c, six characters "same as above" and "principal" as shown in FIG. 2 are predetermined as characters to be converted 23. Character to be converted 2
3 is a designation character that designates the same character as a recorded character in a predetermined field (hereinafter referred to as an index field).

Here, the converted character 23 of "same as above" specifies the same character as the character recorded in the address index field (here, the address field 21a). Furthermore, the character to be converted 23 of "principal" specifies the same character as the character recorded in the first and last name index field (in this case, the first and last name field 22a).

The control unit 15 stores in advance, as format control information, index field information, converted character 23 information, and field identification information for identifying an address or a name field.

Next, the form 20 as shown in FIG. 2 is scanned by the transport scanning section II, and the preprocessing section 12 and character recognition section 13 perform recognition processing on the characters recorded on the form 20. As a result, it is assumed that the recognition results for one form of 20 minutes as shown in FIG. 3 are stored in the post-processing section 14. Post-processing section 1
4 is the recognition result shown in FIG.
The conversion processing for the character to be converted 23 is performed based on the various information outputted from the converter 5 as described above.

That is, when the recognition result in the address fields 21a to 21c specified by the field identification information matches the conversion target character 23 "same as above," the post-processing unit 14 determines that address field to be the conversion target field. Next, the address index field is recognized from the index field information from the control unit 15. The address index field is determined in advance to be the address field 21a in FIG. 2. The post-processing unit 14 recognizes the recorded characters 30 in the address field 21a, which is an address index field, as the same characters as the converted characters 23, and converts each converted character 23 of "same as above" into the recorded characters 30. The post-processing unit 14 also processes the first and last name fields 22a to 22 specified by the field identification information.
If the recognition result in c matches the character to be converted 23, ``principal'', the address field 22c is determined to be the field to be converted. Next, the name index field is recognized from the index field information from the control unit 15. The first and last name index field is determined in advance to be the first and last name field 22a in FIG.

The post-processing unit 14 recognizes the recorded character 31 in the first and last name field 22a, which is the first and last name index field, as the same character as the converted character 23, and converts the converted character 23 of ``1 PERSON'' into the recorded character 31.

In this way, the post-processing section 14 executes the conversion process of the converted characters 23 for 20 minutes of one form, and outputs the final recognition result as shown in FIG. 5 to the control section 15. Control part 1
5 outputs this final recognition result as an answer. Therefore, when a character entry person writes the same character as a recorded character in a predetermined index field, he or she only writes the converted character 23 such as "same as above" or "principal" in the predetermined recording field. That's fine. As a result, for the converted character 23, the same recognition result as when the same character as the recorded character in the index field is recognized is output.

[Effects of the Invention] As described in detail below, according to the present invention, by simply entering a simple and versatile character to be converted, it is possible to convert the same character as the recorded character in the same form specified in advance. It is possible to obtain recognition results for the same character without having to fill in any information. Therefore, the person filling out the form can omit filling in the same characters over and over again, and the burden of filling in the form can be significantly reduced. Furthermore, it is effective in reading a wide range of characters such as kanji, alphanumeric characters, and kana characters without being restricted by the characters recorded on the form.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of an optical character reading device according to an embodiment of the present invention, FIG. 2 is a diagram showing an example of a form according to the embodiment, and FIGS. 3 and 4 are diagrams showing the same embodiment. FIG. 5 is a diagram illustrating an example of a conventional form. to, 20... form, 13... character recognition section, 14
... Post-processing section, 15... Control section. Applicant's Representative Patent Attorney Takehiko Suzue Figure 1 Figure 2 Figure 3 Figure 4 Figure 5

Claims (1)

[Claims] A form having recording fields distinguished in advance for each recording content, a character recognition means for recognizing characters recorded in the recording fields of this form, and a target for instructing conversion into the same recorded characters recorded on the form. A storage means for pre-storing converted character information, index field information indicating a recording field in which the same recorded character indicated by the converted character information is recorded, and field identification information for identifying the recording field; When the recognition result of the character recognition means for the character to be converted recorded in a predetermined recording field of the form is detected based on the memory contents of the means, the character recognition means detects the recognition result of the character to be converted recorded in a predetermined recording field of the form, and the character recognition means detects the recognition result of the character to be converted that is recorded in a predetermined recording field of the form. A character reading device comprising: a control means for outputting a recognition result of the recorded character as a final recognition result of the converted character.