Character recognition method, character recognition apparatus and financial apparatus

• US 9,563,811 B2
• Filed: 07/31/2013
• Issued: 02/07/2017
• Est. Priority Date: 07/31/2012
• Status: Active Grant

• Current Assignee: ATEC AP Co., Ltd. (ATEC T& Co., Ltd.)
• Original Assignee: LG CNS Company Limited (LG Corporation)
• Inventors: Koo, Beum Yong
• Primary Examiner(s): BELLA, MATTHEW C
• Assistant Examiner(s): Shin, Soo
• Application Number:
  US13/955,782
• Publication Number:
  US 20140037181A1
• Time in Patent Office: 1,287 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/15 Cutting or merging image elements, e.g. region growing, watershed or clustering-based techniques
• CPC Class Codes: G06V 30/166 Normalisation of pattern dimensions
• CPC Class Codes: G06V 30/224 of printed characters having additional code marks or containing code marks
• CPC Class Codes: G06V 30/244 using graphical properties, e.g. alphabet type or font

Abstract

A character recognition method for recognizing a character of a medium is provided. A character image of an individual character from a medium is acquired and the character image is read out step by step to determine the character according to a hierarchical structure in which a set of predetermined characters are hierarchically classified into a plurality of groups configured of main groups and sub groups.

Claims

1. A character recognition apparatus, comprising:

a processor configured to;

acquire a character image of an individual character from a medium;

determine the individual character from the character image with a hierarchical structure in which a set of predetermined characters are hierarchically classified into a plurality of groups configured into main groups and sub groups; and

determine whether the determined individual character matches at least one of error-prone characters,

wherein the error-prone characters are previously stored in a memory; and

wherein a predetermined alphabet character printed on a same position of a character of different media is determined from a set of predetermined number characters when the character recognition apparatus performs a recognition process, which acquires a character image of an individual character and determines the individual character, on a plurality of different media repeatedly.

2. The character recognition apparatus of claim 1, wherein any one of the number of character pixels, the number of background pixels, and a ratio of the number of character pixels is compared with the number of background pixels in a predetermined area of the character image with a reference value according to each of hierarchical layers in the hierarchical structure and determines the individual character based on a comparison result.

3. The character recognition apparatus of claim 1, wherein the hierarchical structure classifies the set of predetermined characters into the plurality of groups based on similarity of shapes between characters included in the set of predetermined characters.

4. The character recognition apparatus of claim 1, wherein a group among the plurality of groups is determined in which the character image is included, and the individual character of the character image is determined as one among characters included in the determined group.

5. The character recognition apparatus of claim 1, wherein the hierarchical structure hierarchically classifies numbers 0 to 9 into the plurality of groups configured of main groups and sub groups, and a specific alphabet is determined as a specific number when the character recognizer reads out a character image of the specific alphabet using the hierarchical structure.

6. The character recognition apparatus of claim 1, wherein the hierarchical structure hierarchically classifies numbers 0 to 9 into the plurality of groups configured of main groups and sub groups, and Alphabet E is determined as number 2 when a character image of Alphabet E is read out by the hierarchical structure.

7. The character recognition apparatus of claim 6, wherein the hierarchical structure hierarchically classifies Alphabets A to Z into the plurality of groups configured of main groups and sub groups, and number 8 is determined as Alphabet B when a character image of number 8 is read out by the hierarchical structure.

8. The character recognition apparatus of claim 1, wherein the individual character is determined by at least one template according to kinds of characters included in the set of predetermined characters.

9. The character recognition apparatus of claim 1, wherein the same character position of different media is a position of a last digit of a serial number, and when the predetermined Alphabet character comprises Alphabet E, K or C, the set of predetermined number characters comprises 5, 6, and 2.

10. The character recognition apparatus of claim 1, wherein the same character position of different media is a position of a last digit of a serial number, and when the predetermined Alphabet character comprises Alphabet E, the set of predetermined number characters comprises 2.

11. The character recognition apparatus of claim 1, wherein the same character position of different media is a position of a last digit of a serial number, and when the predetermined Alphabet character comprises Alphabet A, B, D, H, or J, the set of predetermined number characters comprises 3, 4, 8, and 9.

12. The character recognition apparatus of claim 1, wherein the same character position of different media is a position of a last digit of a serial number, and when the predetermined Alphabet character comprises Alphabet L, the set of predetermined number characters comprises 1 and 7.

13. The character recognition apparatus of claim 1, wherein the medium is a first country'"'"'s bill on which a position of a number character of a serial number of the first country'"'"'s bill is determined to be represented by an Alphabet character of a serial number of a second country'"'"'s bill, and

when the character recognition apparatus recognizes the medium, the Alphabet character of the serial number of the second country'"'"'s bill is determined as a number in a set including one or more among numbers 0 to 9.

14. The character recognition apparatus of claim 13, wherein the character image of an Alphabet character of the serial number of the second country'"'"' bill replaced on the position of the number character of the serial number of the first country'"'"'s bill includes a background pixel having a pixel value of 200 or more and a character pixel having a pixel value of 30 or less.

15. The character recognition apparatus of claim 1,

wherein the error-prone character that was previously stored in the memory is extracted from the determined individual character.

16. A financing apparatus, comprising:

a processor configured to;

deposit or withdraw a medium including characters;

acquire a character image of an individual character from the medium, and reading out the character image to determine the characters with a hierarchic structure in which a set of predetermined characters is hierarchically classified into a plurality of groups configured into main groups and sub groups;

control operations of the medium interface and the character recognition apparatus; and

determine whether each of the characters determined matches at least one of error-prone characters,

wherein the error-prone characters are previously stored in a memory; and

wherein a predetermined alphabet character printed on a same position of a character of different media is determined from a set of predetermined number characters when the character recognition apparatus performs recognition process, which acquires a character image of an individual character and determines the individual character, on a plurality of different media repeatedly.

17. The financing apparatus of claim 16,

wherein the error-prone character that was previously stored in the memory is extracted from the determined individual character.