Method and system for recognizing information on a card

• US 9,367,753 B2
• Filed: 05/30/2014
• Issued: 06/14/2016
• Est. Priority Date: 06/03/2013
• Status: Active Grant

• Current Assignee: Alipay.Com Co., Ltd. (Ant Group Co., Ltd.)
• Original Assignee: Alipay.Com Co., Ltd. (Ant Group Co., Ltd.)
• Inventors: Li, Yang, Chen, Guo
• Primary Examiner(s): Niu, Feng
• Application Number:
  US14/292,288
• Publication Number:
  US 20140355883A1
• Time in Patent Office: 746 Days
• Field of Search: None
• US Class Current: 1/1
• CPC Class Codes: G06T 5/73 Deblurring; Sharpening
• 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 30/1444 Selective acquisition, locating or processing of specific regions, e.g. highlighted text, fiducial marks or predetermined fields

Abstract

Embodiments of the present application relate to a method for recognizing information, a system for recognizing information, and a computer program product for recognizing information. A method for recognizing information is provided. The method includes locating a card zone for each frame within a card image frame sequence comprising a plurality of frames, locating an information zone within each card zone, dividing each information zone into at least one character zone, de-blurring a character zone corresponding to a same region across all the frames in the card image frame sequence, and recognizing character string information based on the de-blurred character zone.

Claims

1. A method for recognizing information, comprising:

locating, using one or more processors, a card zone for each frame within a card image frame sequence comprising a plurality of frames;

locating, using the one or more processors, an information zone within each card zone;

dividing, using the one or more processors, each information zone into at least one character zone;

de-blurring, using the one or more processors, a character zone corresponding to a same region across all the frames in the card image frame sequence, wherein the de-blurring of the character zone comprises;

for character zones corresponding to a same region across all the frames in the image frame sequence, calculating a clarity of a first character zone, comprising;

obtaining an intensity value of each pixel in the first character zone based on horizontal and vertical gradient features of the each pixel within the first character zone, comprising;

multiplying a first grayscale value of a first pixel in the first character zone with a corresponding first element of a matrix to obtain a first intensity value of the first pixel;

multiplying a second grayscale value of a second pixel in the first character zone with a corresponding second element of the matrix to obtain a second intensity value of the second pixel, the first element being different from the second element;

determining whether the each pixel is located at an edge of the first character zone; and

in the event that the each pixel is located at the edge of the first character zone, adding a supplemental pixel to the edge of the first character zone, the supplemental pixel having a grayscale value of zero; and

determining the clarity of the first character zone based on the first and second intensity values in the first character zone;

selecting as a candidate character zone, the candidate character zone having a highest clarity among the corresponding character zones; and

recognizing, using the one or more processors, character string information based on the de-blurred character zone.

2. The method as described in claim 1, wherein:

in the event that the clarity of the candidate character zone is higher than or equal to a preset threshold value, regarding the candidate character zone as a selected image of the corresponding character zone for subsequent recognition processing; and

in the event that the clarity of the candidate character zone is lower than the preset threshold value, acquiring a next image frame sequence to be recognized.

3. The method as described in claim 1, further comprising:

prior to the locating of the card zone for each frame within the card image frame sequence;

storing, in a cache, a video stream or a plurality of single-frame images taken of a card face; and

acquiring the image frame sequence from the cache.

4. The method as described in claim 3, wherein the acquiring of the image frame sequence from the cache comprises:

selecting, from the video stream in the cache, various frames in the image frame sequence at intervals of a predetermined number of frames.

5. The method as described in claim 1, further comprising:

verifying whether character string information complies with a predetermined rule for a card;

in the event that the character string information complies with the predetermined rule, displaying the character string information as card information to a user to be used in confirmation; and

in the event that the character string information does not comply with the predetermined rule, acquiring a next image frame sequence to be recognized.

6. The method as described in claim 1, wherein the intensity value of the each pixel relates to each element of the matrix and a greyscale value of the each pixel.

7. A system for recognizing information, comprising:

at least one processor configured to;

locate a card zone for each frame within a card image frame sequence comprising a plurality of frames;

locate an information zone within each card zone;

divide each information zone into at least one character zone;

de-blur a character zone corresponding to a same region across all the frames in the card image frame sequence, wherein the de-blurring of the character zone comprises;

for character zones corresponding to a same region across all the frames in the image frame sequence, calculate a clarity of a first character zone, comprising;

obtain an intensity value of each pixel in the first character zone based on horizontal and vertical gradient features of the each pixel within the first character zone, comprising;

 

multiply a first grayscale value of a first pixel in the first character zone with a corresponding first element of a matrix to obtain a first intensity value of the first pixel;

 

multiply a second grayscale value of a second pixel in the first character zone with a corresponding second element of the matrix to obtain a second intensity value of the second pixel, the first element being different from the second element;

 

determine whether the each pixel is located at an edge of the first character zone; and

 

in the event that the each pixel is located at the edge of the first character zone, add a supplemental pixel to the edge of the first character zone, the supplemental pixel having a grayscale value of zero; and

determine the clarity of the first character zone based on the first and second intensity values in the first character zone;

select as a candidate character zone, the candidate character zone having a highest clarity among the corresponding character zones; and

recognize character string information based on the de-blurred character zone; and

a memory coupled to the at least one processor and configured to provide the at least one processor with instructions.

8. The system as described in claim 7, wherein:

in the event that the clarity of the candidate character zone is higher than or equal to a preset threshold value, regard the candidate character zone as a selected image of the corresponding character zone for subsequent recognition processing; and

in the event that the clarity of the candidate character zone is lower than the preset threshold value, acquire a next image frame sequence to be recognized.

9. The system as described in claim 7, wherein the locating of the card zone for each frame within the card image frame sequence further comprises:

prior to the locating of the card zone for each frame within the card image frame sequence;

store, in a cache, a video stream or a plurality of single-frame images taken of a card face; and

acquire the image frame sequence from the cache.

10. The system as described in claim 9, wherein the acquiring of the image frame sequence from the cache comprises:

select from the video stream in the cache various frames in the image frame sequence at intervals of a predetermined number of frames.

11. The system as described in claim 7, wherein the at least one processor is further configured to:

verify whether character string information complies with a predetermined rule for a card;

in the event that the character string information complies with the predetermined rule, display the character string information as card information to a user to be used in confirmation; and

in the event that the character string information does not comply with the predetermined rule, acquire a next image frame sequence to be recognized.

12. The system as described in claim 7, wherein the intensity value of the each pixel relates to each element of the matrix and a greyscale value of the each pixel.

13. A system for recognizing information, comprising:

a camera configured to acquire a card image frame sequence comprising a plurality of frames;

at least one processor configured to;

locate a card zone for each frame within a card image frame sequence;

locate an information zone within each card zone;

divide each information zone into at least one character zone;

de-blur a character zone corresponding to a same region across all the frames in the card image frame sequence, wherein the de-blurring of the character zone comprises;

for character zones corresponding to a same region across all the frames in the image frame sequence, calculate a clarity of a first character zone, comprising;

obtain an intensity value of each pixel in the first character zone based on horizontal and vertical gradient features of the each pixel within the first character zone, comprising;

 

multiply a first grayscale value of a first pixel in the first character zone with a corresponding first element of a matrix to obtain a first intensity value of the first pixel;

 

multiply a second grayscale value of a second pixel in the first character zone with a corresponding second element of the matrix to obtain a second intensity value of the second pixel, the first element being different from the second element;

 

determine whether the each pixel is located at an edge of the first character zone; and

 

in the event that the each pixel is located at the edge of the first character zone, add a supplemental pixel to the edge of the first character zone, the supplemental pixel having a grayscale value of zero; and

determine the clarity of the first character zone based on the first and second intensity values in the first character zone;

select as a candidate character zone, the candidate character zone having a highest clarity among the corresponding character zones; and

recognize character string information based on the de-blurred character zone; and

a memory coupled to the at least one processor and configured to provide the at least one processor with instructions; and

a display configured to display the recognized character string information.

14. The system as described in claim 13, wherein:

in the event that the clarity of the candidate character zone is higher than or equal to a preset threshold value, regard the candidate character zone as a selected image of the corresponding character zone for subsequent recognition processing; and

in the event that the clarity of the candidate character zone is lower than the preset threshold value, acquire a next image frame sequence to be recognized.

15. The system as described in claim 13, wherein the locating of the card zone for each frame within the card image frame sequence further comprises:

prior to the locating of the card zone for each frame within the card image frame sequence;

store, in a cache, a video stream or a plurality of single-frame images taken of a card face; and

acquire the image frame sequence from the cache.

16. The system as described in claim 15, wherein the acquiring of the image frame sequence from the cache comprises:

select, from the video stream in the cache, various frames in the image frame sequence at intervals of a predetermined number of frames.

17. The system as described in claim 13, wherein the at least one processor is further configured to:

verify whether character string information complies with a predetermined rule for a card;

in the event that the character string information complies with the predetermined rule, display the character string information as card information to a user to be used in confirmation; and

in the event that the character string information does not comply with the predetermined rule, acquire a next image frame sequence to be recognized.

18. The system as described in claim 13, wherein the intensity value of the each pixel relates to each element of the matrix and a greyscale value of the each pixel.

19. A computer program product for recognizing information, the computer program product being embodied in a tangible non-transitory computer readable storage medium and comprising computer instructions for:

locating a card zone for each frame within a card image frame sequence comprising a plurality of frames;

locating an information zone within each card zone;

dividing each information zone into at least one character zone;

de-blurring a character zone corresponding to a same region across all the frames in the card image frame sequence, wherein the de-blurring of the character zone comprises;

for character zones corresponding to a same region across all the frames in the image frame sequence, calculating a clarity of a first character zone, comprising;

obtaining an intensity value of each pixel in the first character zone based on horizontal and vertical gradient features of the each pixel within the first character zone, comprising;

multiplying a first grayscale value of a first pixel in the first character zone with a corresponding first element of a matrix to obtain a first intensity value of the first pixel;

multiplying a second grayscale value of a second pixel in the first character zone with a corresponding second element of the matrix to obtain a second intensity value of the second pixel, the first element being different from the second element;

determining whether the each pixel is located at an edge of the first character zone; and

in the event that the each pixel is located at the edge of the first character zone, adding a supplemental pixel to the edge of the first character zone, the supplemental pixel having a grayscale value of zero; and

determining the clarity of the first character zone based on the first and second intensity values in the first character zone;

selecting as a candidate character zone, the candidate character zone having a highest clarity among the corresponding character zones; and

recognizing character string information based on the de-blurred character zone.