Contextual resolution of character assignments

1. A computer-implemented method of optical character recognition, comprising:

receiving at a computer a plurality of proposed character assignments for at least one instance of a character in an image on which optical character recognition is being performed, each of the character assignments representing a character determined from a character set for the at least one instance of the character by the optical character recognition, the at least one instance of the character residing in one or more multi-character character strings, at least one of the multi-character character strings known to exclude a subset of at least one character from the character set based at least in part on the at least one multi-character string's location in the image; and

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the subset of at least one character excluded from the at least one multi-character character string.

2. The method of claim 1 wherein the subset of characters comprises at least one of alpha characters, numeric characters, punctuation characters, special characters, or a specified subset of characters.

3. The method of claim 1 wherein at least two multi-character character strings exclude different subsets of characters.

4. The method of claim 1 further comprising determining the subset of at least one character excluded by a multi-character character string at least in part from information conveyed by the multi-character character string.

5. The method of claim 1 wherein the character on which optical character recognition is being performed is contained in a field conveying particular information; and

further comprising determining the subset of at least one character excluded by a multi-character character string at least in part from information conveyed by the field.

6. The method of claim 1 further comprising selecting a character assignment for the character from the remaining proposed character assignments.

7. The method of claim 1 wherein eliminating at least one of the proposed character assignments comprises eliminating all but one proposed character assignment; and

the method further comprising selecting a character assignment for the character as the remaining proposed character assignment.

8. The method of claim 7 further comprising associating a geometry of the character with the character assignment for use in subsequent character recognition.

9. The method of claim 1 wherein eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the subset of at least one character excluded from the at least one multi-character character string comprises:

referring to a first multi-character character string known to exclude a first subset of at least one character;

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the first subset of at least one character excluded from the first multi-character character string;

referring to a second multi-character character string known to exclude a second subset of at least one character; and

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignment to the second subset of at least one character excluded from the second multi-character character string.

10. An article comprising a machine-readable medium storing instructions adapted to cause one or more machines to perform operations comprising:

receiving a plurality of proposed character assignments for at least one instance of a character in an image on which optical character recognition is being performed, each of the character assignments representing a character determined from a character set for the at least one instance of the character by the optical character recognition, the at least one instance of the character residing in one or more character strings of multiple characters, at least one of the character strings of multiple characters known to exclude a subset of at least one character from the character set based at least in part on the at least one character string of multiple characters' location in the image; and

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the subset of at least one character excluded from the at least one character string of multiple characters.

11. The article of claim 10 wherein the specified subset of characters comprises at least one of alpha characters, numeric characters, punctuation characters, or special characters.

12. The article of claim 10 wherein the instructions are further adapted to cause one or more machines to perform operations comprising:

determining the subset of at least one character excluded by a character string of multiple characters at least in part from information conveyed by the character string of multiple characters.

13. The article of claim 10 wherein the character on which optical character recognition is being performed is contained in a field conveying particular information; and

wherein the instructions are further adapted to cause one or more machines to perform operations comprising determining the subset of at least one character excluded by a character string of multiple characters at least in part from information conveyed by the field.

14. The article of claim 10 wherein the instructions are further adapted to cause one or more machines to perform operations comprising selecting a character assignment for the character from the remaining proposed character assignments.

15. The article of claim 14 wherein the instructions are further adapted to cause one or more machines to perform operations comprising associating a geometry of the character with the character assignment for use in subsequent character recognition.

16. The article of claim 10 wherein eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the subset of at least one character excluded from the at least one character string of multiple characters comprises:

referring to a first character string of multiple characters known to exclude a first subset of at least one character;

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the first subset of at least one character excluded from the first character string of multiple characters;

referring to a second character string of multiple characters known to exclude a second subset of at least one character; and

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the first subset of at least one character excluded from the second character string of multiple characters.

17. A system for optical character recognition including a processor configured to perform operations comprising:

receiving a plurality of proposed character assignments for at least one instance of a character in an image on which optical character recognition is being performed, each of the character assignments representing a character determined from a character set for the at least one instance of the character by the optical character recognition, the at least one instance of the character residing in one or more character strings of a plurality of characters, at least one of the character strings of a plurality of characters known to exclude a subset of at least one character from the character set based at least in part on the at least one character string of a plurality of characters' location in the image; and

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the subset of at least one character excluded from the at least one character string of a plurality of characters.

18. The system of claim 17 wherein the specified subset of characters comprises at least one of alpha characters, numeric characters, punctuation characters, or special characters.

19. The system of claim 17 wherein the processor is further configured to perform operations comprising determining the subset of at least one character excluded by a character string of a plurality of characters at least in part from information conveyed by the character string of a plurality of characters.

20. The system of claim 17 wherein the character on which optical character recognition is being performed is contained in a field conveying particular information; and

wherein the processor is further configured to perform operations comprising determining the subset of at least one character excluded by a character string of a plurality of characters at least in part from information conveyed by the field.

21. The system of claim 17 wherein the processor is further configured to perform operations comprising selecting a character assignment for the character from the remaining proposed character assignments.

22. The system of claim 21 wherein the processor is further configured to perform operations comprising associating a geometry of the character with the character assignment for use in subsequent character recognition.

23. The system of claim 17 wherein eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the subset of at least one character excluded from the at least one character string of a plurality of characters comprises:

referring to a first character string of a plurality of characters known to exclude a first subset of at least one character;

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the first subset of at least one character excluded from the first character string of a plurality of characters;

referring to a second character string of a plurality of characters excluding known to exclude a second subset of at least one character; and

eliminating at least one of the proposed character assignments by comparing the at least one of the proposed character assignments to the second subset of at least one character excluded from the second character string of a plurality of characters.

24. A computer-implemented method of optical character recognition, comprising:

receiving at a computer a plurality of proposed character assignments for at least one instance of a character in an image on which optical character recognition is being performed, the at least one instance of the character residing on an image, each character assignment representing a character determined from a character set for the at least one instance of the character by the optical character recognition;

referring to a multiple-character reference character string containing the at least one instance of the character or another instance of the character and the reference character string known to exclude a subset of one or more characters of the character set based at least in part on the reference character string's location in the image; and

eliminating at least one of the plurality of proposed character assignments by comparing the at least one of the proposed character assignments to the subset of one or more characters known to be excluded by the reference character string.

25. The method of claim 24 further comprising:

referring to a second multiple-character reference character string containing another instance of the character and the second reference character string known to exclude a second subset of one or more characters of the character set based at least in part on the second reference character string's location in the image; and

eliminating at least one of the plurality of proposed character assignments by comparing the at least one of the proposed character assignments to the second subset of one or more characters known to be excluded by the second reference character string.

26. The method of claim 24 wherein the multiple-character reference character string resides on a second image.

27. The method of claim 24 wherein eliminating at least one of the plurality of proposed character assignments comprises eliminating all but one of the plurality of proposed character assignments; and

further comprising selecting a character assignment as the remaining proposed character assignment.

28. The method of claim 27 further comprising associating a geometry of the character with the character assignment for determining subsequent character assignments.

29. The method of claim 1 wherein the image comprises a specified format comprising a plurality of fields, and further wherein the at least one multi-character string's location in the image used to determine the excluded subset of at least one character from the character set comprises the at least one multi-character string's location in a particular one of the plurality of fields.

30. The method of claim 1 wherein the location of the at least one multi-character string in the image used to determine the excluded subset of at least one character from the character set comprises the location of the at least one multi-character string relative to at least one other multi-character string in the image.

31. The method of claim 1 wherein the image comprises a digitized document.

32. The article of claim 10 wherein the image comprises a specified format comprising a plurality of fields, and further wherein the at least one character string of multiple characters' location in the image used to determine the excluded subset of at least one character from the character set comprises the at least one character string of multiple characters' location in a particular one of the plurality of fields.

33. The article of claim 10 wherein the location of the at least one character string of multiple characters in the image used to determine the excluded subset of at least one character from the character set comprises the location of the at least one character string of multiple characters relative to at least one other character string of multiple characters in the image.

34. The article of claim 10 wherein the image comprises a digitized document.

35. The system of claim 17 wherein the image comprises a specified format comprising a plurality of fields, and further wherein the at least one character string of a plurality of characters' location in the image used to determine the excluded subset of at least one character from the character set comprises the at least one character string of a plurality of characters' location in a particular one of the plurality of fields.

36. The system of claim 17 wherein the location of the at least one character string of a plurality of characters in the image used to determine the excluded subset of at least one character from the character set comprises the location of the at least one character string of a plurality of characters relative to at least one other character string of a plurality of characters in the image.

37. The system of claim 17 wherein the image comprises a digitized document.

38. The method of claim 24 wherein the image comprises a specified format comprising a plurality of fields, and further wherein the reference character string's location in the image used to determine the excluded subset of at least one character from the character set comprises the reference character string's location in a particular one of the plurality of fields.

39. The method of claim 24 wherein the location of the reference character string in the image used to determine the excluded subset of at least one character from the character set comprises the location reference character string relative to at least one other reference character string.

40. The method of claim 24 wherein the image comprises a digitized document.