Handwriting recognition by word separation into sillouette bar codes and other feature extraction

• US 20010033694A1
• Filed: 01/19/2001
• Published: 10/25/2001
• Est. Priority Date: 01/19/2000
• Status: Active Grant

• Current Assignee: California Institute of Technology
• Original Assignee: California Institute of Technology
• Inventors: Woods, Donald J., Chen, Joseph, Goodman, Rodney M., Keaton, Patricia A.
• Granted Patent: US 6,917,708 B2
• Time in Patent Office: Days
• US Class Current: 382/229
• CPC Class Codes: G06V 30/2264 using word shape

Abstract

A method of automatically recognizing text. The text is divided into whole words which are each recognize. Each whole word is characterized according to its silhouette. The silhouette is characterized by features in the silhouette such as upwardly extending “polls” and downwardly extending “holes”. The silhouette may also be characterized by its first syllable blends. Numbers are assigned to each of the different characteristics, and numbers may also be assigned based on analysis of a database of different kinds of cursive words. Recognition may be automatically carry out prior recognizing system which recognizes in this way.

Claims

1. A method of recognizing handwriting, comprising:

obtaining a sample of handwriting;

segmenting said sample into separate handwritten words; and

attempting to recognize a whole handwritten word without attempting to recognize any individual letter of the whole handwritten word.

2. A method as in

claim 1

, wherein said recognizing comprises determining a silhouette of the word, and matching said silhouette to one of a plurality of reference silhouettes.

3. A method as in

claim 2

, been said recognizing comprises determining features of the silhouette.

4. A method as in

claim 3

, wherein said features of the silhouette includes high-profile features, and low profile features, and locations of said high-profile features and said low-profile features.

5. A method as in

claim 3

, wherein said determining features comprises determining prime features.

6. A method as in

claim 5

, further comprising super enclosing said prime features to form hybrid features.

7. A method as in

claim 6

, further comprising sorting said features by first syllable blends.

8. A method as in

claim 1

, wherein said attempting comprises categorizing said whole hidden handwritten word according to its overall silhouette.

9. A method as in

claim 8

, wherein said categorizing comprises categorizing positions of features in said handwritten word, and categorizing first syllable blends of said handwritten word.

10. A method as in

claim 1

, wherein said sample of handwriting includes family names.

11. A method as in

claim 10

, further comprising forming a list of a plurality of family names, and forming silhouette information about said plurality of family names, and comparing said separate handwritten words to said plurality of family names.

12. A method as in

claim 11

, wherein said comparing comprises forming silhouette information, and comparing said silhouette information into said silhouette information about said plurality of family names.

13. A method as in

claim 12

, wherein said silhouette information includes information about the presence of high and low parts in the written word and the position of those high and low parts.

14. A method as in

claim 12

wherein said silhouette information includes first syllable blends in the word.

15. A method, comprising:

analyzing a sample of handwriting by analyzing a whole word of said sample at any one time, said analyzing comprising forming information indicative of a silhouette of said whole word, and comparing said information with a database of information about other silhouettes.

16. A method as in

claim 15

, wherein said database of information comprises a database of information obtained from a list of possible words.

17. A method as in

claim 16

wherein said words are family names, and said list of possible words is a telephone book.

18. A method as in

claim 15

, wherein said silhouette information includes information indicative of high parts in the word and low parts in the word, and positions of said high parts and low parts in the word.

19. A method as in

claim 15

, wherein said silhouette information includes information about first syllable blends in the word.

20. A method as in

claim 18

, wherein said silhouette information also includes information about first syllable blends in the word.

21. A method as in

claim 15

, wherein said silhouette information includes prime profiles indicative of specified features, and concatenated profiles indicative of combinations at specified features.

22. A method as in

claim 18

, wherein each of a plurality of silhouette information'"'"'s is provided with a number.

23. A method as in

claim 18

, wherein each feature is assigned a number.