Method for character recognition

• US 8,180,160 B2
• Filed: 11/29/2007
• Issued: 05/15/2012
• Est. Priority Date: 12/01/2006
• Status: Expired due to Fees

• Current Assignee: Cerence Inc., Cerence Operating Company (Cerence Inc.), Robert Bosch GmbH, Robert Bosch LLC (Robert Bosch GmbH), Zi Decuma AB
• Original Assignee: Zi Decuma AB
• Inventors: Sternby, Jakob
• Primary Examiner(s): MARIAM, DANIEL G
• Application Number:
  US11/947,197
• Publication Number:
  US 20080130996A1
• Time in Patent Office: 1,629 Days
• Field of Search: 382181-189, 382/173, 382/224, 382/179
• US Class Current: 382/224
• CPC Class Codes: G06V 30/2272 with lexical matching
• CPC Class Codes: G06V 30/347 Sampling; Contour coding; Stroke extraction

Abstract

The present invention generally describes a method for classifying a line segment of a handwritten line into a reference feature set, wherein said handwritten line comprises one or several curves representing a plurality of symbols. First, sample data representing said handwritten line is received. Next, a sample line segment in said received sample data is identified by detecting a sample line segment start point (SLSSP) and a sample line segment end point (SLSEP). Then, a sample feature set of said identified sample line segment is determined. Finally, the determined sample feature set is matched to a reference feature set among a plurality of reference feature sets.

Claims

1. A method for classifying a line segment of a handwritten line into a reference feature set, wherein said handwritten line comprises one or several curves representing a plurality of symbols, said method comprising:

receiving sample data representing said handwritten line;

identifying a sample line segment in said received sample data by detecting a sample line segment start point (SLSSP) and a sample line segment end point (SLSEP);

determining a sample feature set of said identified sample line segment, wherein said determination of said sample feature set further comprises;

determining a first angle between a gradient of said line segment in said SLSSP and a segment specific reference line intersecting said SLSSP and said SLSEP;

associating said first angle to said sample feature set;

determining a second angle between a gradient of said line segment in said SLSEP and a segment specific reference line intersecting said SLSSP and said SLSEP;

associating said second angle to said sample feature set;

determining a general angle between the segment specific reference line intersecting said SLSSP and said SLSEP and a reference line corresponding to a writing direction; and

associating said general angle to said sample feature set; and

matching said determined sample feature set to a reference feature set among a plurality of reference feature sets,

wherein the method is performed using a computer or processor.

2. The method according to claim 1, wherein said steps of receiving, identifying, determining, and matching are performed successively.

3. The method according to claim 1, wherein said plurality of reference feature sets are predetermined.

4. The method according to claim 1, wherein said determination of said sample feature set further comprises:

determining a length between said SLSSP and said SLSEP; and

associating said length to said sample feature set.

5. The method according to claim 4, wherein said determination of said length further comprises:

generating a normalized length of said length by taking prior determined lengths into account.

6. The method according to claim 1, wherein said determination of said sample feature set further comprises:

determining a type curve matching said line segment; and

associating said type curve to said sample feature set.

7. The method according to claim 1, wherein said reference feature set comprises a minimum value and a maximum value for each feature.

8. The method according to claim 1, wherein said sample data comprises coordinate data.

9. A hardware-implemented module for classifying a line segment of a handwritten line into one reference line segment of a reference line segment set, wherein said handwritten line comprises one or several curves representing a plurality of symbols, said module comprising:

a receiver configured to receive data representing said handwritten line;

an identifier configured to identify a sample line segment in said received data by detecting a sample line segment start point (SLSSP) and a sample line segment end point (SLSEP);

a determinator configured to determine a sample feature set for said identified sample line segment, wherein said determinator further comprises;

an angle determinator configured to determine an angle between said line segment and a reference line intersecting said SLSSP and said SLSEP;

an associater configured to associate said determined angle to said sample feature set;

a general angle determinator configured to determine a general angle between a segment specific reference line intersecting said SLSSP and said SLSEP and a reference line corresponding to a writing direction;

an associater configured to associate said determined general angle to said sample feature set; and

a matcher configured to match said determined sample feature set to a reference feature set among a plurality of reference feature sets; and

a transmitter configured to output said matching reference feature set.

10. The hardware-implemented module according to claim 9, wherein said determinator further comprises:

a length determinator configured to determine a length between said SLSSP and said SLSEP; and

an associater configured to associate said determined length to said sample feature set.

11. The hardware-implemented module according to claim 10, wherein said determinator further comprises:

a storage medium comprising a number of prior determined lengths;

a length transformer configured to normalize said length to a normalized length by taking said prior determined lengths comprised within said storage medium into account; and

an associater configured to associate said determined normalized length to said sample feature set.

12. The hardware-implemented module according to claim 9, wherein said determinator further comprises:

a type curve determinator configured to determine a type curve matching said line segment; and

an associater configured to associate said determined type curve to said sample feature set.

13. An apparatus comprising:

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

a module according to claim 9 configured to receive said data from said touch sensitive area and to output reference feature sets;

a symbol matcher configured to match said reference feature sets into a symbol set corresponding to a symbol; and

a display configured to present said symbol.

14. The apparatus according to claim 13, wherein said pen movement capturing device is a touch sensitive area.

15. The apparatus according to claim 13, further comprising a symbol set database comprising a number of reference line segment combinations and their associated symbol sets.

16. A computer readable memory device having stored thereon instructions that are executable by a computer to classify a line segment of a handwritten line into a reference feature set, wherein the handwritten line comprises one or several curves representing a plurality of symbols, the instructions causing the computer to (a) receive sample data representing the handwritten line, (b) identify a sample line segment in the received sample data by detecting a sample line segment start point (SLSSP) and a sample line segment end point (SLSEP), (c) determine a sample feature set of the identified sample line segment, determine a first angle between a gradient of said line segment in said SLSSP and a segment specific reference line intersecting said SLSSP and said SLSEP, associate said first angle to said sample feature set, determine a second angle between a gradient of said line segment in said SLSEP and a segment specific reference line intersecting said SLSSP and said SLSEP, associate said second angle to said sample feature set, determine a general angle between the segment specific reference line intersecting said SLSSP and said SLSEP and a reference line corresponding to a writing direction, and associate said general angle to said sample feature set, and (d) match the determined sample feature set to a reference feature set among a plurality of reference feature sets.

17. A method for classifying a line segment of a handwritten line into a reference feature set, wherein said handwritten line comprises one or several curves representing a plurality of symbols, said method comprising:

receiving sample data representing said handwritten line;

identifying a sample line segment in said received sample data by detecting a sample line segment start point (SLSSP) and a sample line segment end point (SLSEP);

determining a sample feature set of said identified sample line segment, wherein said determination of said sample feature set further comprises;

determining a first angle between a gradient of said line segment in said SLSSP and a segment specific reference line intersecting said SLSSP and said SLSEP;

associating said first angle to said sample feature set;

determining a second angle between a gradient of said line segment in said SLSEP and a segment specific reference line intersecting said SLSSP and said SLSEP;

associating said second angle to said sample feature set;

determining a general angle between the segment specific reference line intersecting said SLSSP and said SLSEP and a reference line corresponding to a writing direction;

associating said general angle to said sample feature set;

determining a type curve matching said line segment; and

associating said type curve to said sample feature set; and

matching said determined sample feature set to a reference feature set among a plurality of reference feature sets,

wherein the method is performed using a computer or processor.

18. The method according to claim 17, wherein said steps of receiving, identifying, determining, and matching are performed successively, wherein said plurality of reference feature sets are predetermined.

19. The method according to claim 17, wherein said determination of said sample feature set further comprises:

determining a length between said SLSSP and said SLSEP; and

associating said length to said sample feature set.

20. The method according to claim 19, wherein said determination of said length further comprises:

generating a normalized length of said length by taking prior determined lengths into account.

21. A hardware-implemented module for classifying a line segment of a handwritten line into one reference line segment of a reference line segment set, wherein said handwritten line comprises one or several curves representing a plurality of symbols, said module comprising:

a receiver configured to receive data representing said handwritten line;

an identifier configured to identify a sample line segment in said received data by detecting a sample line segment start point (SLSSP) and a sample line segment end point (SLSEP);

a determinator configured to determine a sample feature set for said identified sample line segment, wherein said determinator further comprises;

an angle determinator configured to determine an angle between said line segment and a reference line intersecting said SLSSP and said SLSEP;

an associater configured to associate said determined angle to said sample feature set;

a general angle determinator configured to determine a general angle between a segment specific reference line intersecting said SLSSP and said SLSEP and a reference line corresponding to a writing direction;

an associater configured to associate said determined general angle to said sample feature set;

a type curve determinator configured to determine a type curve matching said line segment; and

an associater configured to associate said determined type curve to said sample feature set;

a matcher configured to match said determined sample feature set to a reference feature set among a plurality of reference feature sets; and

a transmitter configured to output said matching reference feature set.

22. The hardware-implemented module according to claim 21, wherein said determinator further comprises:

a length determinator configured to determine a length between said SLSSP and said SLSEP; and

an associater configured to associate said determined length to said sample feature set.

23. The hardware-implemented module according to claim 22, wherein said determinator further comprises:

a storage medium comprising a number of prior determined lengths;

a length transformer configured to normalize said length to a normalized length by taking said prior determined lengths comprised within said storage medium into account; and

an associater configured to associate said determined normalized length to said sample feature set.