Method and apparatus for character recognition, and computer-readable recording medium with a program making a computer execute the method recorded therein

• US 6,320,983 B1
• Filed: 11/03/1998
• Issued: 11/20/2001
• Est. Priority Date: 03/27/1998
• Status: Expired due to Term

• Current Assignee: Fujitsu Limited
• Original Assignee: Fujitsu Limited
• Inventors: Kobara, Katsutoshi, Chiba, Koichi, Nagano, Yoshihiro, Matsuno, Hideki, Eguchi, Shinichi
• Primary Examiner(s): Au, Amelia M.
• Assistant Examiner(s): Dastouri, Mehrdad
• Application Number:
  US09/185,604
• Time in Patent Office: 1,113 Days
• Field of Search: 382/137, 382/138, 382/140, 382/187, 382/186, 382/190, 382/189, 382/209, 382/229, 382/188, 705/45
• US Class Current: 382/186
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/1444 Selective acquisition, locating or processing of specific regions, e.g. highlighted text, fiducial marks or predetermined fields

Abstract

A character recognition apparatus, in which a type of each of accounts, a type of each of marks, and characteristic data indicating a string of characters recognized when any of the marks is written in each account are previously registered for each group of the types of accounts in correlation to each other in an account-type database; a mark is estimated according to a result of verification between the account-type database and a result of ordinary character recognition according to a character recognizing program; and a type of account having the estimated mark is determined as a selected account name according to a result of recognizing characters in accounts other than the account having the mark.

Claims

1. A character recognition apparatus for recognizing characters based on an image data obtained by optically reading a form having a plurality of items forming one group printed thereon each expressed by a character string for selection comprising:

a database in which a plurality types of marks are previously prepared and an item name for each of the items, a type of each of the marks and a result of character recognition when each of the mark is written in each of the item are previously registered in correlation to each other;

a verifying unit for verifying the result of character recognition of the image data to those registered in said database;

an estimation unit for estimating a mark from the plurality types of marks based on the result of verification by said verifying unit; and

a selected item determining unit for discriminating an item that corresponds and items that do not correspond to the mark estimated by said estimating unit while referring to said database, obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition unit, and determining the item name corresponding to the obtained item as a selected item name.

2. A character recognition apparatus according to claim 1;

wherein in said database are registered items without a mark correlated thereto previously.

3. A character recognition apparatus according to claim 1;

wherein said selected item determining unit determines that no item to be selected when said estimating unit estimates that all of the items are without any mark.

4. A character recognition apparatus according to claim 1;

wherein the plurality types of marks include a selection mark indicating a selected item and an exclusion mark indicating exclusion of an item from those to be selected, and said selected item determining unit determines, when the exclusion mark is estimated by said estimating unit, an item name for an item other than those each having the exclusion mark as a selected item name.

5. A character recognition apparatus according to claim 1;

wherein said estimating unit extracts a plurality of candidates for selection according to a result of verification by said verifying unit and estimates a mark that occurs most frequently in the plurality of candidates as a final candidate.

6. A character recognition apparatus for recognizing characters based on an image data obtained by optically reading a form having a plurality of items forming one group printed thereon each expressed by a character string or selection comprising:

a database with an arrangement of the plurality of items previously decided according to a pre-specified order on the form, in which a plurality types of marks are previously prepared and an item name for each of the items, a type of each of the marks, result of character recognition when each of the marks is written in each of the items and information for a position where each item is arranged are previously registered in correlation to each other therein;

a verifying unit for verifying the result of character recognition of the image data to those registered in said database;

an estimation unit for estimating a mark from the plurality types of marks based on the result of verification by said verifying unit; and

a selected item determining unit for extracting an item name from the plurality of item names by referring to the database according to the position of the mark estimated by said estimating unit and determining the extracted item name as a selected item name by obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition unit.

7. A character recognition apparatus according to claim 6;

wherein in said database are registered items without a mark correlated thereto previously.

8. A character recognition apparatus according to claim 6;

wherein said selected item determining unit determines that no item to be selected when said estimating unit estimates that all of the items are without any mark.

9. A character recognition apparatus according to claim 6;

wherein the plurality types of marks include a selection mark indicating a selected item and an exclusion mark indicating exclusion of an item from those to be selected, and said selected item determining unit determines, when the exclusion mark is estimated by said estimating unit, an item name for an item other than those each having the exclusion mark as a selected item name.

10. A character recognition apparatus according to claim 6;

wherein said estimating unit extracts a plurality of candidates for selection according to a result of verification by said verifying unit and estimates a mark that occurs most frequently in the plurality of candidates as a final candidate.

11. A character recognition apparatus with a plurality of groups each comprising a plurality of items each expressed by a character string as a choice as a unit for executing character recognition according to image data obtained by optically reading a form with any of the plurality of groups printed thereon;

a database in which a plurality types of marks are previously prepared and an item name for each of the item names, a type of each of the marks and a result of character recognition when each of the marks is written in each of the items are previously registered in correlation to each other for each of the groups;

a verifying unit for verifying the result of character recognition of the image data to those registered in said database for each item;

an estimation unit for estimating a mark from the plurality types of marks based on the result of verification by said verifying unit; and

a selected item determining unit for discriminating an item that has been marked with the mark estimated by said estimating unit from the items that are not marked using this mark by referring to the database, preparing item names for the items that are not marked by the mark estimated by said estimating unit by using the result of character recognition, identifying an appropriate group from the plurality of groups according to the prepared item names and determining the item name of the item that has been marked with the mark estimated by said estimating unit in the identified group as a selected item name from the prepared item names and obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition unit.

12. A character recognition apparatus according to claim 11;

wherein in said database are registered items without a mark correlated thereto previously.

13. A character recognition apparatus according to claim 11;

wherein said selected item determining unit determines that no item to be selected when said estimating unit estimates that all of the items are without any mark.

14. A character recognition apparatus according to claim 11;

wherein the plurality types of marks include a selection mark indicating a selected item and an exclusion mark indicating exclusion of an item from those to be selected, and said selected item determining unit determines, when the exclusion mark is estimated by said estimating unit, an item name for an item other than those each having the exclusion mark as a selected item name.

15. A character recognition apparatus according to claim 11;

wherein said estimating unit extracts a plurality of candidates for selection according to a result of verification by said verifying unit and estimates a mark that occurs most frequently in the plurality of candidates as a final candidate.

16. A character recognition apparatus with a plurality of groups each comprising a plurality of items each expressed by a character string as a choice as a unit for executing character recognition according to image data obtained optically reading a form with any of the plurality of groups printed thereon;

a database in which a plurality types of marks are previously prepared and an item name for each of the item names, a type of each of the marks and a result of character recognition when each of the marks is written in each the items are previously registered in correlation to each other for each of the groups;

a verifying unit for verifying the result of character recognition of the image data to those registered in said database for each item;

a estimating unit for estimating a mark from the plurality types of marks based on the result of verification by said verifying unit; and

a selected item determining unit for extracting an item name from the plurality of item names by referring to the database according to the position of a mark estimated by said estimating unit and determining the extracted item name as a selected item name and obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

17. A character recognition apparatus according to claim 16;

wherein in said database are registered items without a mark correlated thereto previously.

18. A character recognition apparatus according to claim 16;

wherein said selected item determining unit determines that no item to be selected when said estimating unit estimates that all of the items are without any mark.

19. A character recognition apparatus according to claim 16;

wherein the plurality types of marks include a selection mark indicating a selected item and an exclusion mark indicating exclusion of an item from those to be selected, and said selected item determining unit determines, when the exclusion mark is estimated by said estimating unit, an item name for an item other than those each having the exclusion mark as a selected item name.

20. A character recognition apparatus according to claim 16;

wherein said estimating unit extracts a plurality of candidates for selection according to a result of verification by said verifying unit and estimates a mark that occurs most frequently in the plurality of candidates as a final candidate.

21. A character recognition method applied to a character recognition apparatus for recognizing characters based on an image data obtained by optically reading a form having a plurality of items forming one group printed thereon each expressed by a character string for selection as a group comprising the steps of:

a first step of providing a database, in which the plurality types of marks are previously prepared and an item name for each of the items, a type of each of the marks and a result of character recognition when each of the mark is written in each of the item are previously registered in correlation to each other, and verifying the result of character recognition of the image data to those registered in the database for each item;

a second step of estimating a mark from the plurality types of marks based on the result of verification in said first step; and

a third step of discriminating an item that has been marked with the mark estimated in said second step from the items that are not marked using this mark by referring to the database, preparing an item name for the items that are not marked by the mark estimated in said second step by using the result of character recognition and determining the item name of the item that has been marked with the mark estimated in said second step as a selected item name from the prepared item names by obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

22. A character recognition method applied to a character recognition apparatus for recognizing characters based on an image data obtained by optically reading a form having a plurality of items forming one group printed thereon each expressed by a character string for selection comprising the steps of:

a first step of providing a database, in which a plurality types of marks are previously prepared with arrangement of the plurality of items previously decided according to a prespecified order on the form, an item name for each of the items, a type of each of the marks, a result of character recognition when each of the mark is written in each of the item and information on a position where each item is arranged are previously registered in correlation to each other, and verifying the result of character recognition of the image data to those registered in the database for each item;

a second step of estimating a mark from the plurality types of marks based on the result of verification in said first step; and

a third step of extracting an item name from a position of the mark estimated in said second step among the plurality of items by referring to said database and determining the extracted item name as a selected item name by obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

23. A character recognition method applied to a character recognition apparatus with a plurality of groups each comprising a plurality of items each expressed by a character string as a choice as a unit for recognizing characters based on an image data obtained by optically reading a form with one of the plurality of groups printed thereon comprising the steps of:

a first step of providing a database in which a plurality types of marks are previously prepared and an item name for each of the items, a type of each of the marks and a result of character recognition when each of the mark is written in each of the items are previously registered in correlation to each other for each group, and verifying the result of character recognition of the image data to those registered in the database for each item;

a second step of estimating a mark from the plurality types of marks based on the result of verification in said first step; and

a third step of discriminating an item that has been marked with the mark estimated in said second step from the items that are not marked using this mark by referring to the database, preparing an item name for the items that are not marked by the mark estimated in said second step by using the result of character recognition, identifying an appropriate group from the plurality of groups according to the prepared item names and determining the item name of the item that has been marked with the mark estimated in said second step as a selected item name from the prepared item names and obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

24. A character recognition method applied to a character recognition apparatus with a plurality of groups each comprising a plurality of items each expressed by a character string as a choice as a unit for recognizing character based on an image data obtained by optically reading a form with one of the plurality of groups printed thereon comprising the steps of:

a first step of providing a database in which a plurality types of marks are previously prepared and an item name for each of the items, a type of each of the marks and a result of character recognition when each of the marks is written in each of the items are previously registered in correlation to each other for each group, and verifying the result of character recognition of the image data to those registered in the database for each item;

a second step of estimating a mark from the plurality types of marks based on the result of verification in said first step; and

a third step of extracting an item name among the plurality of items from a position of the mark estimated in said second step by referring to said database and determining the extracted item name as a selected item name and obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

25. A computer-readable recording medium with a program recorded therein for making a computer execute the character recognition method applied to a character recognition apparatus for recognizing characters based on an image data obtained by optically reading a form with a group comprising a plurality of items each expressed by a character string as a choice printed thereon;

said character recognition apparatus comprising a database, in which a plurality types of marks are previously prepared therein and an item name for each of the items, a type of each of the marks and a result of character recognition when each of the marks is written in each of the items are previously registered in correlation to each other;

said method comprising;

a first step of verifying the result of character recognition of the image data to those registered in the database for each item;

a second step of estimating a mark from the plurality types of marks based on the result of verification in said first step; and

a third step of discriminating an item having the mark estimated in said second step among the plurality of items from those other than the item by referring to the database with item names prepared for items other than that having the mark estimated in said second step according to a result of character recognition for the image data and determining the item name of the item that has been marked with a mark estimated in said second step as a selected item name from the prepared item names and obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

26. A computer-readable recording medium with a program recorded therein for making a computer execute the character recognition method applied to a character recognition apparatus for recognizing characters based on an image data obtained by optically reading a form with a group comprising a plurality of items each expressed by a character string as a choice printed thereon;

said character recognition apparatus comprising a database with arrangement of the plurality of items previously decided according to a prespecified order on the form, in which a plurality types of marks are previously prepared therein and also an item name for each of the items, a type of each of the marks, a result of character recognition when each of the marks is written in each of the items, and information for a position where each item is arranged are previously registered in previously registered in correlation to each other

said method comprising;

a first step of verifying a result of character recognition of the image data to those in the database for each item;

a second step of estimating a mark from the plurality types of marks based on a result of verification in said first step; and

a third step of extracting an item name from a position of the mark estimated in said second step among the plurality of items by referring to said database and determining the extracted item name as a selected item name and obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

27. A computer-readable recording medium with a program recorded therein for making a computer execute the character recognition method applied to a character recognition apparatus with a plurality of groups each comprising a plurality of items each expressed by a character string as a choice as a unit for recognizing characters based on image data obtained by optically reading a form with any of the plurality of groups printed thereon;

said character recognition apparatus comprising a database, in which a plurality types of marks are previously prepared and also an item name for each of the items, a type of each of the marks, and a result of character recognition when each of the marks is written in each of the items are previously registered in correlation to each other for each group,

said method comprising;

a first step of verifying a result of character recognition of the image data to those in the database for each item;

a second step of estimating a mark from the plurality types of marks based on a result of verification in said first step; and

a third step of discriminating an item that has been marked with the mark estimated in said second step from the items that are not marked using this mark by referring to the database, preparing an item name for the items that are not marked by the mark estimated in said second step by using the result of character recognition, identifying an appropriate group from the plurality of groups according to the previously prepared item names and determining the item name of the item that has been marked with the mark estimated in said second step as a selected item name from the prepared item names and obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

28. A computer-readable recording medium with a program recorded therein for making a computer execute the character recognition method applied to a character recognition apparatus with a plurality of groups each comprising a plurality of items each expressed by a character string as a choice as a unit for recognizing characters based on an image data obtained by optically reading a form with any of the plurality of groups printed thereon;

said character recognition apparatus comprising a database, in which a plurality types of marks are previously prepared and also an item name for each of the items, a type of each of the marks, and a result of character recognition when each of the marks is written in each of the items are previously registered in correlation to each other for each group,

said method comprising;

a first step of verifying the result of character recognition of the image data to those in the database for each item;

a second step of estimating a mark from the plurality types of marks based on a result of verification in said first step; and

a third step of extracting an item name from a position of the mark estimated in said second step among the plurality of items by referring to the database and determining the extracted item name as a selected item name and obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition.

29. A character recognition apparatus for recognizing characters based on an image data obtained by optically reading a form having a plurality of items forming one group printed thereon each expressed by a character string for selection comprising:

a scanner for reading a form having a plurality of items forming a group printed thereon;

a character recognition unit for recognizing characters based on the image data obtained by said scanner;

a database in which a plurality of types of marks are previously prepared and an item name for each of the items, a type of each of the marks and a result of character recognition when each of the mark is written in each of the item are previously registered in correlation to each other;

a verifying unit for verifying the result of character recognition of the image data to those registered in said database;

an estimation unit for estimating a mark from the plurality types of marks based on the result of verification by said verifying unit; and

a selected item determining unit for discriminating an item that corresponds and items that do not correspond to the mark estimated by said estimating unit while referring to said database, obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition unit, and determining the item name corresponding to the obtained item as a selected item name.

30. A character recognition apparatus for recognizing characters based on an image data obtained by optically reading a form having a plurality of items forming one group printed thereon each expressed by a character string for selection comprising:

a scanner for reading a form having a plurality of items forming a group printed thereon;

a character recognition unit for recognizing characters based on the image data obtained by said scanner;

a database with an arrangement of the plurality of items previously decided according to a pre-specified order on the form, in which a plurality types of marks are previously prepared and an item name for each of the items, a type of each of the marks, result of character recognition when each of the marks is written in each of the items and information for a position where each item is arranged are previously registered in correlation to each other therein;

a verifying unit for verifying the result of character recognition of the image data to those registered in said database;

an estimation unit for estimating a mark from the plurality types of marks based on the result of verification by said verifying unit; and

a selected item determining unit for extracting an item name from the plurality of item names by referring to the database according to the position of the mark estimated by said estimating unit and determining the extracted item name as a selected item name by obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition unit.

31. A character recognition apparatus with a plurality of groups each comprising a plurality of items each expressed by a character string as a choice as a unit for executing character recognition according to image data obtained by optically reading a form with any of the plurality of groups printed thereon;

a scanner for reading a form having a plurality of items forming a group printed thereon;

a character recognition unit for recognizing characters based on the image data obtained by said scanner

a database in which a plurality types of marks are previously prepared and an item name for each of the item names, a type of each of the marks and a result of character recognition when each of the marks is written in each of the items are previously registered in correlation to each other for each of the groups;

a verifying unit for verifying the result of character recognition of the image data to those registered in said database for each item;

an estimation unit for estimating a mark from the plurality types of marks based on the result of verification by said verifying unit; and

a selected item determining unit for discriminating an item that has been marked with the mark estimated by said estimating unit from the items that are not marked using this mark by referring to the database, preparing item names for the items that are not marked by the mark estimated by said estimating unit by using the result of character recognition, identifying an appropriate group from the plurality of groups according to the prepared item names and determining the item name of the item that has been marked with the mark estimated by said estimating unit in the identified group as a selected item name from the prepared item names by obtaining the item corresponding to the estimated mark based on character recognition of items that do not correspond to the estimated mark by said character recognition unit.