SYSTEM AND METHOD FOR RECOGNIZING TEXT INFORMATION IN OBJECT

• US 20130004076A1
• Filed: 02/07/2012
• Published: 01/03/2013
• Est. Priority Date: 06/29/2011
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: You, Kisun, KOO, HYUNG-IL, Cho, Hyun-Mook
• Granted Patent: US 9,418,304 B2
• Time in Patent Office: Days
• US Class Current: 382/176
• CPC Class Codes: G06F 18/254 of classification results, e.g. of results related to same input data
• CPC Class Codes: G06V 10/809 of classification results, e.g. where the classifiers operate on the same input data
• CPC Class Codes: G06V 20/62 Text, e.g. of license plates, overlay texts or captions on TV images
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/142 using hand-held instruments; Constructional details of the instruments

Abstract

A method for recognizing a text block in an object is disclosed. The text block includes a set of characters. A plurality of images of the object are captured and received. The object in the received images is then identified by extracting a pattern in one of the object images and comparing the extracted pattern with predetermined patterns. Further, a boundary of the object in each of the object images is detected and verified based on predetermined size information of the identified object. Text blocks in the object images are identified based on predetermined location information of the identified object. Interim sets of characters in the identified text blocks are generated based on format information of the identified object. Based on the interim sets of characters, a set of characters in the text block in the object is determined.

Claims

1. A method for recognizing a text block in an object, the text block having a set of characters, the method comprising:

receiving a plurality of images of the object, each image of the object including the text block;

identifying the text blocks in the plurality of images of the object; and

determining the set of characters based on the identified text blocks in the plurality of images of the object.

2. The method of claim 1, wherein determining the set of characters includes:

generating an interim set of characters for each of the identified text blocks; and

inferring the set of characters based on the interim sets of characters.

3. The method of claim 2, wherein generating the interim sets of characters includes generating the interim sets of characters based on a holistic method.

4. The method of claim 2, wherein generating the interim sets of characters includes generating the interim sets of characters based on a segmentation-based method.

5. The method of claim 2, wherein inferring the set of characters includes inferring the set of characters based on a confidence level of the interim sets of characters.

6. The method of claim 5, wherein a character in the set of characters is inferred based on a number of occurrences in the corresponding position in the interim sets of characters.

7. The method of claim 6, wherein the character in the set of characters is inferred when the number of occurrences exceeds a predetermined threshold value.

8. The method of claim 2, wherein generating the interim set of characters includes:

receiving format information on the set of characters in the text block; and

generating the interim set of characters based on the received format information.

9. The method of claim 1, wherein identifying the text blocks includes:

detecting a boundary of the object in each of the plurality of images of the object;

verifying the detected boundary based on predetermined size information of the object; and

identifying the text blocks based on predetermined location information of the text block of the object.

10. The method of claim 9, wherein verifying the detected boundary includes receiving a new image of the object if the detected boundary does not match the predetermined size information.

11. The method of claim 1, further comprising determining a type of the object based on a predetermined pattern extracted from a portion of a reference object image, the predetermined pattern being unique to the type of the object.

12. The method of claim 1, wherein the plurality of images of the object include video frames of the object.

13. The method of claim 1, further comprising displaying the determined set of characters.

25. The device of claim 13, wherein the plurality of images of the object include video frames of the object.

26. The device of claim 13, further comprising a display configured to display the determined set of characters.

14. A device for recognizing a text block in an object, the text block having a set of characters, the device comprising:

an image capturing unit configured to capture a plurality of images of the object, each image of the object including the text block;

a masking unit configured to identify the text blocks in the plurality of images of the object; and

a temporal filter configured to determine the set of characters based on the identified text blocks in the plurality of images of the object.

15. The device of claim 14, further comprising an OCR (optical character reader) engine configured to generate an interim set of characters for each of the identified text blocks,

wherein the set of characters is inferred based on the interim sets of characters by the temporal filter.

16. The device of claim 15, wherein the OCR engine is further configured to generate the interim set of characters based on a holistic method.

17. The device of claim 15, wherein the OCR engine is further configured to generate the interim set of characters based on a segmentation-based method.

18. The device of claim 15, wherein the temporal filter is further configured to infer the set of characters based on a confidence level of the interim sets of characters.

19. The device of claim 18, wherein a character in the set of characters is inferred based on a number of occurrences in the corresponding position in the interim sets of characters.

20. The device of claim 19, wherein the character in the set of characters is inferred when the number of characters exceeds a predetermined threshold value.

21. The device of claim 15, wherein the OCR engine is further configured to receive format information on the set of characters in the text block and generate the interim set of characters based on the received format information.

22. The device of claim 14, further comprising a boundary detector configured to detect a boundary of the object in each of the plurality of images of the object and verify the detected boundary based on predetermined size information of the object,

wherein the masking unit is further configured to identify the text blocks based on predetermined location information of the text block of the object.

23. The device of claim 22, wherein the image capturing unit is further configured to receive a new image of the object if the detected boundary does not match the predetermined size information.

24. The device of claim 14, further comprising an object recognition unit configured to determine a type of the object based on a predetermined pattern extracted from a portion of a reference object image, the predetermined pattern being unique to the type of the object.

27. A non-transitory computer-readable medium comprising instructions for recognizing a text block in an object, the instructions causing a processor to perform of the operations of:

receiving a plurality of images of the object, each image of the object including the text block;

identifying the text blocks in the plurality of images of the object; and

determining a set of characters in the text block based on the identified text blocks in the plurality of images of the object.

28. The medium of claim 27, wherein determining the set of characters includes:

generating an interim set of characters for each of the identified text blocks; and

inferring the set of characters based on the interim sets of characters.

29. The medium of claim 28, wherein generating the interim sets of characters includes generating the interim sets of characters based on a holistic method.

30. The medium of claim 28, wherein generating the interim sets of characters includes generating the interim sets of characters based on a segmentation-based method.

31. The medium of claim 28, wherein inferring the set of characters includes inferring the set of characters based on a confidence level of the interim sets of characters.

32. The medium of claim 31, wherein a character in the set of characters is inferred based on a number of occurrences in the corresponding position in the interim sets of characters.

33. The medium of claim 32, wherein the character in the set of characters is inferred when the number of occurrences exceeds a predetermined threshold value.

34. The medium of claim 28, wherein generating the interim set of characters includes:

receiving format information on the set of characters in the text block; and

generating the interim set of characters based on the received format information.

35. The medium of claim 27, wherein identifying the text blocks includes:

detecting a boundary of the object in each of the plurality of images of the object;

verifying the detected boundary based on predetermined size information of the object; and

identifying the text blocks based on predetermined location information of the text block of the object.

36. The medium of claim 35, wherein verifying the detected boundary includes receiving a new image of the object if the detected boundary does not match the predetermined size information.

37. The medium of claim 27, wherein the operations further comprise determining a type of the object based on a predetermined pattern extracted from a portion of a reference object image, the predetermined pattern being unique to the type of the object.

38. The medium of claim 27, wherein the plurality of images of the object include video frames of the object.

39. The medium of claim 27, wherein the operations further comprise displaying the determined set of characters.

40. An apparatus for recognizing a text block in an object, the text block having a set of characters, the apparatus comprising:

means for capturing a plurality of images of the object, each image of the object including the text block;

means for identifying the text blocks in the plurality of images of the object; and

means for determining the set of characters based on the identified text blocks in the plurality of images of the object.

41. The apparatus of claim 40, further comprising means for generating an interim set of characters for each of the identified text blocks,

wherein the set of characters is inferred based on the interim sets of characters by the determining means.

42. The apparatus of claim 41, wherein the determining means infers the set of characters based on a confidence level of the interim sets of characters.

43. The apparatus of claim 42, wherein a character in the set of characters is inferred based on a number of occurrences in the corresponding position in the interim sets of characters.

44. The apparatus of claim 43, wherein the character in the set of characters is inferred when the number of characters exceeds a predetermined threshold value.

45. The apparatus of claim 41, wherein the generating means receives format information on the set of characters in the text block and generates the interim set of characters based on the received format information.

46. The apparatus of claim 40, further comprising:

means for detecting a boundary of the object in each of the plurality of images of the object; and

means for verifying the detected boundary based on predetermined size information of the object,

wherein the identifying means identifies the text blocks based on predetermined location information of the text block of the object.

47. The apparatus of claim 46, wherein the capturing means receives a new image of the object if the detected boundary does not match the predetermined size information.

48. The apparatus of claim 40, further comprising means for determining a type of the object based on a predetermined pattern extracted from a portion of a reference object image, the predetermined pattern being unique to the type of the object.