Address reader, sorting machine such as a mail thing and character string recognition method

• US 6,470,091 B2
• Filed: 04/02/2001
• Issued: 10/22/2002
• Est. Priority Date: 02/10/1998
• Status: Expired due to Fees

• Current Assignee: Hitachi-Omron Terminal Solutions Corporation (Hitachi, Ltd.)
• Original Assignee: Hitachi, Ltd.
• Inventors: Watanabe, Shigeru, Ueda, Yoh, Kagehiro, Tatsuhiko, Koga, Masashi, Fujisawa, Hiromichi
• Primary Examiner(s): Johns, Andrew W.
• Assistant Examiner(s): NAKHJAVAN, SHERVIN K
• Application Number:
  US09/822,254
• Publication Number:
  US 20010021261A1
• Time in Patent Office: 568 Days
• Field of Search: 382/101, 382/102, 382/175, 382/176, 382/177, 382/179, 382/186, 382/188, 382/189, 382/198, 209/3.1, 209/509, 209/584, 209/900
• US Class Current: 382/101
• CPC Class Codes: G06V 30/416 Extracting the logical structure, e.g. chapters, sections or page numbers; Identifying elements of the document, e.g. authors
• CPC Class Codes: Y10S 209/90 Sorting flat-type mail

Abstract

An address reader method and apparatus for recognizing a receiver address on a surface of mail. An image of the surface of the mail is input and segmented into at least one character string candidate. At least one address area candidate is extracted from the image based upon the segmented character string candidate. One address area candidate extracted from the image is selected as a receiver address by analyzing each address area candidate based on predetermined position information indicating a usual position of a receiver address area, character direction information indicating a character direction of a character string appropriate for the predetermined position information, and key character string information indicating a character string most likely to exist in a receiver address. Characters in character strings of the selected address area candidate are recognized as a receiver address.

Claims

1. An address reader method of recognizing a receiver address on a surface of mail, said address reader method comprising:

inputting an image of the surface of mail;

segmenting said image into at least one character string candidate;

extracting at least one address area candidate based on said at least one character string candidate;

selecting one of said at least one address area candidate as a receiver address area of the mail by analyzing each of said at least one address area candidate based on predetermined position information indicating a usual position of a receiver address area, character direction information indicating a character direction of a character string appropriate for said predetermined position information, and key character string information including a character string having plural characters most likely to exist in a receiver address; and

recognizing characters in character strings of the selected address area candidate as a receiver address.

2. An address reader method according to claim 1, wherein said selecting comprises:

determining a direction of each character string included in each address area candidate.

3. An address reader method according to claim 1, wherein said selecting comprises:

comparing each character string included in each address area candidate to key character strings, each having characters most likely to exist in a receiver address.

4. An address reader method according claim 1, wherein said selecting comprises:

comparing each character string included in each address area candidate to key character strings which indicate cities most likely to be included in a receiver address area.

5. An address reader method according to claim 1, wherein said position information includes information indicating a usual position of a key character string in a receiver address area.

6. An address reader method according to claim 1, wherein said selecting comprises:

testing each address area candidate to determine whether said address area candidate is an address area candidate most likely to contain a receiver address.

7. An address reader method according to claim 6, wherein said testing comprises:

making an assumption that character strings contained in said address area candidate are in a particular character direction and analyzing said character strings based on the assumption.

8. An address reader method according to claim 7, wherein said making an assumption comprises:

detecting characters of a character string based on reference character information.

9. An address reader method according to claim 1, wherein said selecting comprises:

analyzing, based on position relation information, relative positions between address area candidates which include key character strings so as to distinguish between an address area candidate containing a receiver address and an address area candidate containing a sender address.

10. An address reader method according to claim 1, wherein said selecting comprises:

referring to sender information to determine whether each address area candidate indicated as containing a key character string contains sender information.

11. An address reader apparatus for recognizing a receiver address on a surface of mail, said address reader apparatus comprising:

an image input apparatus which inputs an image of the surface of mail;

an address processing apparatus which segments said image into at least one character string candidate, extracts at least one address area candidate based on said at least one character string candidate, and selects one of said at least one address area candidate as a receiver address area of the mail by analyzing each of said at least one address area candidate based on predetermined position information indicating a usual position of a receiver address area, character direction information indicating a character direction of a character string appropriate for said predetermined position information, and key character string information including a character string having plural characters most likely to exist in a receiver address; and

an address recognition module which recognizes characters in character strings of the selected address area candidate as a receiver address.

12. An address reader apparatus according to claim 11, wherein said address processing apparatus determines a direction of each character string included in each address area candidate.

13. An address reader apparatus according to claim 11, wherein said address processing apparatus compares each character string included in each address area candidate to key character strings, each having characters most likely to exist in a receiver address.

14. An address reader apparatus according claim 11, wherein said address processing apparatus compares each character string included in each address area candidate to key character strings which indicate cities most likely to be included in a receiver address area.

15. An address reader apparatus according to claim 11, wherein said position information includes information indicating a usual position of a key character string in a receiver address area.

16. An address reader apparatus according to claim 11, wherein said address processing apparatus tests each address area candidate to determine whether said address area candidate is an address area candidate most likely to contain a receiver address.

17. An address reader apparatus according to claim 16, wherein said address processing apparatus when performing said testing makes an assumption that character strings contained in said address area candidate are in a particular character direction and analyzing said character strings based on the assumption.

18. An address reader apparatus according to claim 17, wherein said address processing apparatus when performing said making an assumption detects characters of a character string based on reference character information.

19. An address reader apparatus according to claim 11, wherein said address processing apparatus analyzes, based on position relation information, relative positions between address area candidates which include key character strings so as to distinguish between an address area candidate containing a receiver address and an address area candidate containing a sender address.

20. An address reader apparatus according to claim 11, wherein said address processing apparatus refers to sender information to determine whether each address area candidate indicated as containing a key character string contains sender information.

21. A mail sorting machine for sorting mail, comprising:

a scanner which scans a surface of the mail and outputs an image of the surface of the mail based on the scan;

an address reader apparatus for recognizing a receiver address on a surface of mail, said address reader apparatus comprises;

an image input apparatus which inputs said image of the surface of the mail from said scanner,

an address processing apparatus which segments said image into at least one character string candidate, extracts at least one address area candidate based on said at least one character string candidate, and selects one of said at least one address area candidate as a receiver address area of the mail by analyzing each of said at least one address area candidate based on predetermined position information indicating a usual position of a receiver address area, character direction information indicating a character direction of a character string appropriate for said predetermined position information, and key character string information including a character string having plural characters most likely to exist in a receiver address, and

an address recognition module which recognizes characters in character strings of the selected address area candidate as a receiver address; and

a sorter which sorts the mail based on the receiver address recognized by said address recognition module.

22. A mail sorting machine according to claim 21, wherein said address processing apparatus determines a direction of each character string included in each address area candidate.

23. A mail sorting machine according to claim 21, wherein said address processing apparatus compares each character string included in each address area candidate to key character strings, each having characters most likely to exist in a receiver address.

24. A mail sorting machine according claim 21, wherein said address processing apparatus compares each character string included in each address area candidate to key character strings which indicate cities most likely to be included in a receiver address area.

25. A mail sorting machine according to claim 21, wherein said position information includes information indicating a usual position of a key character string in a receiver address area.

26. A mail sorting machine according to claim 21, wherein said address processing apparatus tests each address area candidate to determine whether said address area candidate is an address area candidate most likely to contain a receiver address.

27. A mail sorting machine according to claim 26, wherein said address processing apparatus when performing said testing makes an assumption that character strings contained in said address area candidate are in a particular character direction and analyzing said character strings based on the assumption.

28. A mail sorting machine according to claim 27, wherein said address processing apparatus when performing said making an assumption detects characters of a character string based on reference character information.

29. A mail sorting machine according to claim 21, wherein said address processing apparatus analyzes, based on position relation information, relative positions between address area candidates which include key character strings so as to distinguish between an address area candidate containing a receiver address and an address area candidate containing a sender address.

30. A mail sorting machine according to claim 21, wherein said address processing apparatus refers to sender information to determine whether each address area candidate indicated as containing a key character string contains sender information.