Method For Additive Character Recognition And An Apparatus Thereof

• US 20070206859A1
• Filed: 02/28/2007
• Published: 09/06/2007
• Est. Priority Date: 03/01/2006
• Status: Active Grant

• Current Assignee: Cerence Operating Company (Cerence Inc.)
• Original Assignee: Zi Decuma AB
• Inventors: Sternby, Jakob
• Granted Patent: US 7,865,016 B2
• Time in Patent Office: Days
• US Class Current: 382/186
• CPC Class Codes: G06V 30/1423 the instrument generating sequences of position coordinates corresponding to handwriting preprocessing or recognising digital ink G06V30/32
• CPC Class Codes: G06V 30/347 Sampling; Contour coding; Stroke extraction
• CPC Class Codes: G06V 30/36 Matching; Classification

Abstract

A method for recognition of a handwritten pattern comprising one or more curves is presented. The method comprises a step of receiving sample data representing the handwritten pattern. The method further comprises a step of segmenting the handwritten pattern by detecting segmentation points on each curve, and by dividing the handwritten pattern into segments. Further, the method comprises a step of comparing the handwritten pattern to templates wherein the comparing comprises a step of normalizing said segments according to a scheme which is independent of the templates to which the segments are to be compared, and a step of determining matching measures for selecting at least one sequence of templates representing a recognintion candidate of the handwritten pattern.

Claims

1. A method for recognition of a handwritten pattern comprising one or more curves, said method comprising:

receiving sample data representing the handwritten pattern;

segmenting the handwritten pattern by detecting segmentation points on each curve, and by dividing the handwritten pattern into segments; and

comparing the handwritten pattern to templates wherein the comparing comprises normalizing said segments according to a scheme which is independent of the templates to which the segments are to be compared, and determining matching measures for selecting at least one sequence of templates representing a recognintion candidate of the handwritten pattern.

2. The method according to claim 1, wherein said matching measures comprise segmental matching measures comparing segmental features of the handwritten pattern to segmental features of the templates.

3. The method according to claim 1, wherein said matching measures comprise connective matching measures comparing connective features between segments in the handwritten pattern to connective features of templates.

4. The method according to claim 2, further comprising a step of compensating for translation, angle or length differences between the segments such that the segmental features are relative within each possible segment.

5. The method according to claim 3, further comprising a step of compensating for translation, angle or length differences between the segments such that the connective features are relative between the adjacent segments.

6. The method according to claim 2, wherein said segmental features comprise a segmental distance between two segments.

7. The method according to claim 2, wherein said segmental features comprise a distance component between two pairs of attached segments.

8. The method according to claim 3, wherein said connective features comprise a distance component for non-connected segments.

9. The method according to claim 3, wherein said connective features comprise a distance component for a connection between two segments.

10. The method according to claims 6, 7, 8 or 9 wherein said step of determining matching measures utilizes an operator in order to determine the connection of templates that are to be used as a model for comparison with the connections between segments.

11. The method according to claim 6 or 9, wherein said operator is a linear function of the segmental distance between two segments and the distance component for a connection between two segments.

12. The method according to claim 10, wherein said operator is a linear function of the segmental distance between two segments and the distance component for a connection between two segments.

13. The method according to claim 1, further comprising a step of detecting the segmentation points as local extreme points which are below a predetermined threshold.

14. The method according to claim 13, further comprising a step of parameterizing each segment by the Dijkstra Curve Maximization strategy with three intermittent points.

15. The method according to claim 1, wherein said step of comparing utilizes point-to-curve matching.

16. The method according to claim 1, further comprising a step of associating an output weight to normalized segmental and connective features.

36. A computer program arranged to perform the method according to any of the claims 1-16 when downloaded into and run on a computational device.

17. A module for recognition of a handwritten pattern comprising one or more curves, said module comprising:

a receiver configured to receive sample data representing the handwritten pattern;

a segmentation point detector configured to detect segmentation points on each curve;

a divider configured to divide the handwritten pattern into segments;

a normalizer configured to normalize said segments according to a scheme which is independent of the templates to which the segments are to be compared;

a determinator configured to determine matching measures for selecting at least one sequence of templates representing a recognition candidate of the handwritten pattern; and

a transmitter configured to output said matching templates.

18. The module according to claim 17, wherein said determinator is configured to determine segmental matching measures.

19. The module according to claim 17, wherein said determinator is configured to determine connective matching measures.

20. The module according to claim 18, further comprising a compensator configured to compensate for translation, angle or length differences between the segments such that the segmental features are relative within each possible segment.

21. The module according to claim 19, further comprising a compensator configured to compensate for translation, angle or length differences between the segments such that the connective features are relative between the adjacent segments.

22. The module according to claim 18, wherein said determinator is configured to determine a segmental distance between two segments.

23. The module according to claim 18, wherein said determinator is configured to determine a distance component between two pairs of attached segments.

24. The module according to claim 19, wherein said determinator is configured to determine a distance component for non-connected segments.

25. The module according to claim 19, wherein said determinator is configured to determine a distance component for a connection between two segments.

26. The module according to claims 22, 23, 24 or 25 wherein said determinator utilizes an operator in order to determine the connection of templates that are to be used as a model for comparison with the connections between segments.

27. The module according to claim 22 or 25, wherein said operator is a linear function of the segmental distance between two segments and the distance component for a connection between two segments.

28. The module according to claim 26, wherein said operator is a linear function of the segmental distance between two segments and the distance component for a connection between two segments.

29. The module according to claim 17, wherein said segmental point detector is configured to detect the segmentation points as local extreme points which are below a predetermined threshold.

30. The module according to claim 26, wherein said segmental point detector is configured to parameterize each segment by the Dijkstra Curve Maximization strategy with three intermittent points.

31. The module according to claim 17, wherein said determinator is configured to utilize point-to-curve matching.

32. The module according to claim 17, further comprising an associator configured to associate an output weight to every normalized segmental and connective feature.

33. An apparatus comprising:

a pen movement capturing device configured to receive data representing a handwritten pattern;

a module according to claim 17 configured to receive said data from said pen movement capturing device and to output matching templates;

a symbol matcher configured to match said templates into symbols; and

a display configured to present said symbols.

34. An apparatus according to claim 33, wherein said pen movement capturing device is a touch sensitive surface.

35. An apparatus according to any of claim 33 or 34, further comprising a symbol set database comprising a number of reference template combinations and their associated symbols.