Adaptive partial character recognition

• US 8,594,431 B2
• Filed: 09/14/2012
• Issued: 11/26/2013
• Est. Priority Date: 12/30/2008
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Ben-Horesh, Ami, Geva, Amir, Walach, Eugeniusz
• Primary Examiner(s): ENTEZARI, MICHELLE M
• Application Number:
  US13/617,106
• Publication Number:
  US 20130011067A1
• Time in Patent Office: 438 Days
• Field of Search: 382/189, 382/321
• US Class Current: 382/187
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/155 Removing patterns interfering with the pattern to be recognised, such as ruled lines or underlines
• CPC Class Codes: G06V 30/1902 Shifting or otherwise transforming the patterns to accommodate for positional errors
• CPC Class Codes: G06V 30/248 involving plural approaches, e.g. verification by template match; Resolving confusion among similar patterns, e.g. "O" versus "Q" G06V30/242 takes precedence

Abstract

A method and system for recognizing a character affected by a noise or an obstruction is disclosed. After receiving an image with characters, a character being affected by a noise or an obstruction is determined. Then, areas in the character where the noise or obstruction affected are precisely located. Templates representing every possible character in the image are updated by removing equivalent areas to the areas in the character being affected by the noise or obstruction. Then, the character is classified in a template among the updated templates by finding the template having the highest number of matching pixels with the character.

Claims

1. A system for identifying a character with a noise or an obstruction, said system comprising:

a memory device;

a processor coupled to said memory device, wherein said processor is configured to perform steps of;

receiving an image that includes characters;

calculating an average stroke feature of said characters by using one of;

(1) obtaining a stroke width of each character in said image, summing the obtained stroke width of each character, and dividing said summation by the number of said characters in said image;

or (2) dividing the image into multiple areas;

obtaining a stroke width of each area in said image, adding the obtained stroke width of each area, and dividing a result of said addition by the number of said areas;

measuring a stroke feature of a first character in said image;

comparing the measured stroke feature of said first character against the calculated average stroke feature of said characters;

determining, based on the comparing, whether said measured stroke feature of said first character does not conform to said calculated average stroke feature of said characters;

detecting a noise or an obstruction in said first character in response to determining that said measured stroke feature of said first character does not conform to said calculated average stroke feature of said characters;

receiving one or more templates of a plurality of characters;

updating said one or more received templates based on said detected noise or said detected obstruction; and

classifying said character in one of said one or more updated templates.

2. The system according to claim 1, wherein said stroke feature represents one or more of:

a movement of a pen, a movement of pencil, a movement of a stylus, a movement of a digital writing tablet, a movement of a digital touch screen interface.

3. The system according to claim 1, wherein said stroke feature of said first character includes a stroke width of said first character.

4. The system according to claim 1, wherein said processor is further configured to perform a step of:

measuring statistics of adjacent characters.

5. The system according to claim 4, wherein said statistics includes one or more of:

a font height, an ascent value, and a descent value.

6. The system according to claim 1, wherein in order to detect said noise or said obstruction, said processor is further configured to perform steps of:

measuring a stroke feature of adjacent characters of said first character; and

finding an area of said first character in which said measured stroke feature of said first character does not conform to said measured stroke feature of said adjacent characters.

7. The system according to claim 1, wherein in order to perform the detecting said noise or said obstruction, said processor is further configured to perform steps of:

dividing said character into parts;

analyzing each part to determine whether said each part is affected by said noise or said obstruction by comparing a stroke feature of said each part to said calculated average stroke feature of said characters;

upon determining a particular part is affected by said noise or said obstruction, dividing said particular part into subparts; and

analyzing each subpart to determine whether said each subpart is affected by said noise or said obstruction by comparing a stroke feature of said each subpart to said calculated average stroke feature of said characters.

8. The system according to claim 7, in order to update said one or more templates, wherein the processor is further configured to perform steps of:

removing areas in said one or more templates, said removed area being equivalent to areas, in said first character, which are affected by said detected noise or said obstruction.

9. The system according to claim 8, wherein said processor is further configured to perform steps of:

removing said affected areas in said first character.

10. The system according to claim 1, wherein in order to classify said character, said processor is further configured to perform steps of:

counting a number of matching first-type pixels in a template and said first character;

counting a number of matching second-type pixels in said template and said first character;

setting pixels in at least one area in said first character as don'"'"'t care pixels, said at least one area being affected by said noise or said obstruction;

counting a number of said don'"'"'t care pixels in said first character;

calculating a value of (said number of matching first-type pixels+said number of matching second-type pixels-said number of don'"'"'t care pixels) for each pair of each template and said first character;

selecting a pair of a template and said first character that has a highest calculated value among the calculated values of pairs of each template and said first character; and

identifying said first character as a letter, a number or a symbol that said template in said selected pair represents.

11. A method for identifying a character with a noise or an obstruction, said method comprising:

receiving an image that includes characters;

calculating an average stroke feature of said characters by using one of;

(1) obtaining a stroke width of each character in said image, summing the obtained stroke width of each character, and dividing said summation by the number of said characters in said image;

or (2) dividing the image into multiple areas;

obtaining a stroke width of each area in said image, adding the obtained stroke width of each area, and dividing a result of said addition by the number of said areas;

measuring a stroke feature of a first character in said image;

comparing the measured stroke feature of said first character against the calculated average stroke feature of said characters;

determining, based on the comparing, whether said measured stroke feature of said first character does not conform to said calculated average stroke feature of said characters;

detecting a noise or an obstruction in said first character in response to determining that said measured stroke feature of said first character does not conform to said calculated average stroke feature of said characters;

receiving one or more templates of a plurality of characters;

updating said one or more received templates based on said detected noise or said detected obstruction; and

classifying said character in one of said one or more updated templates.

12. The method according to claim 11, wherein said stroke feature represents one or more of:

a movement of a pen, a movement of pencil, a movement of a stylus, a movement of a digital writing tablet, a movement of a digital touch screen interface.

13. The method according to claim 11, wherein said stroke feature of said first character includes a stroke width of said first character.

14. The method according to claim 11, further comprising:

measuring statistics of said first characters.

15. The method according to claim 14, wherein said statistics includes one or more of:

a font height, an ascent value, and a descent value.

16. The method according to claim 11, wherein said detecting includes:

measuring a stroke feature of adjacent characters of said first character; and

finding an area of said first character in which said measured stroke feature of said first character does not conform to said measured stroke feature of said adjacent characters.

17. The method according to claim 16, wherein said detecting comprises:

dividing said character into parts;

analyzing each part to determine whether said each part is affected by said noise or said obstruction by comparing a stroke feature of said each part to said calculated average stroke feature of said characters;

upon determining a particular part is affected by said noise or said obstruction, dividing said particular part into subparts; and

analyzing each subpart to determine whether said each subpart is affected by said noise or said obstruction by comparing a stroke feature of said each subpart to said calculated average stroke feature of said characters.

18. The method according to claim 17, wherein said updating includes a step of:

removing areas in said one or more templates, said removed area being equivalent to areas, in said first character, which are affected by said detected noise or said obstruction.

19. The method according to claim 18, further comprising:

removing said affected areas in said first character.

20. The method according to claim 11, wherein said classifying includes steps of:

counting a number of matching first-type pixels in a template and said first character;

counting a number of matching second-type pixels in said template and said first character;

setting pixels in at least one area in said first character as don'"'"'t care pixels, said at least one area being affected by said noise or said obstruction;

counting a number of said don'"'"'t care pixels in said first character;

calculating a value of (said number of matching first-type pixels+said number of matching second-type pixels-said number of don'"'"'t care pixels) for each pair of each template and said first character;

selecting a pair of a template and said first character that has a highest calculated value among the calculated values of pairs of each template and said first character; and

identifying said first character as a letter, a number or a symbol that said template in said selected pair represents.

21. A computer program product for identifying a character with a noise or an obstruction, the computer program product comprising a non-transitory storage medium readable by a processing circuit and storing instructions run by the processing circuit for performing a method, said method steps comprising:

receiving an image that includes characters;

calculating an average stroke feature of said characters by using one of;

(1) obtaining a stroke width of each character in said image, summing the obtained stroke width of each character, and dividing said summation by the number of said characters in said image;

or (2) dividing the image into multiple areas;

obtaining a stroke width of each area in said image, adding the obtained stroke width of each area, and dividing a result of said addition by the number of said areas;

measuring a stroke feature of a first character in said image;

comparing the measured stroke feature of said first character against the calculated average stroke feature of said characters;

determining, based on the comparing, whether said measured stroke feature of said first character does not conform to said calculated average stroke feature of said characters;

detecting a noise or an obstruction in said first character in response to determining that said measured stroke feature of said first character does not conform to said calculated average stroke feature of said characters;

receiving one or more templates of a plurality of characters;

updating said one or more received templates based on said detected noise or said detected obstruction; and

classifying said character in one of said one or more updated templates.

22. The computer program product according to claim 21, wherein said classifying includes steps of:

counting a number of matching first-type pixels in a template and said first character;

counting a number of matching second-type pixels in said template and said first character;

setting pixels in at least one area in said first character as don'"'"'t care pixels, said at least one area being affected by said noise or said obstruction;

counting a number of said don'"'"'t care pixels in said first character;

calculating a value of (said number of matching first-type pixels+said number of matching second-type pixels-said number of don'"'"'t care pixels) for each pair of each template and said first character;

selecting a pair of a template and said first character that has a highest calculated value among the calculated values of pairs of each template and said first character; and

identifying said first character as a letter, a number or a symbol that said template in said selected pair represents.