Optical character recognition method and apparatus

• US 5,131,053 A
• Filed: 08/10/1988
• Issued: 07/14/1992
• Est. Priority Date: 08/10/1988
• Status: Expired due to Term

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Caere Corp. (Mexico) (Microsoft Corporation)
• Inventors: George, David, Knight, Jeremy, Higgins, Bryan, Dilworth, John, Bernzott, Philip
• Primary Examiner(s): Moore, David K.
• Assistant Examiner(s): Cammarata, Michael
• Application Number:
  US07/230,847
• Time in Patent Office: 1,434 Days
• Field of Search: 382/38, 382/22, 382/40, 382/9, 358/462
• US Class Current: 382/176
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/146 Aligning or centring of the image pick-up or image-field
• CPC Class Codes: G06V 30/414 Extracting the geometrical structure, e.g. layout tree; Block segmentation, e.g. bounding boxes for graphics or text

Abstract

A system for recognition of characters on a medium. The system includes a scanner for scanning a medium such as a page of printed text and graphics and producing a bit-mapped representation of the page. The bit-mapped representation of the page is then stored in a memory means such as the memory of a computer system. A processor processes the bit-mapped image to produce an output comprising coded character representation of the text on the page. The present invention discloses parsing a page to allow for production of the output characters in a logical sequence, a combination of feature detection methods and template matching methods for recognition of characters and a number of methods for feature detection such as use of statistical data and polygon fitting.

Claims

1. A method for performing optical character recognition of textual material on a medium, said medium having at least one array element, said method identifying said array element as consisting essentially of text or as comprising non-text, said identification of said array element as consisting essentially of text, comprising the steps of;

(a) determining whether the number of on-bits in a first predetermined number of bytes of said array element is greater than or equal to a second predetermined number and less than or equal to a third predetermined number; and

(b) determining whether the number of phase changes from on-bits to off-bits in said first predetermined number of bytes of said array element is greater than or equal to a fourth predetermined number and less than or equal to a fifth predetermined number.

2. The method as recited by claim 1, wherein said first predetermined number is four, said second predetermined number is 2, said third predetermined number is 12, said fourth predetermined number is 8 and said fifth predetermined number is 12.

3. A system for optically scanning a medium, said medium having thereon an unknown character, said system comprising:

scanning means for scanning said medium, said scanning means providing as output a bit-mapped image of said medium;

a first memory means coupled with said scanning means for storing said bit-mapped image;

a second memory means for storing a plurality of bit-mapped templates representing known characters;

processing means coupled with said first memory means and said second memory means, said processing means including means for parsing said bit-mapped image of said medium and providing as output a bit-mapped representation of said unknown character, means for generating polygon representations descriptive of the shape of said unknown character, means for generating statistical information concerning said polygon representations and means for analyzing said statistical information to identify said unknown character.

4. The system as recited in claim 3, wherein said processing means further includes means for identifying said unknown character by comparing said bit-mapped representation of said unknown character with each associated one of said plurality of templates of known characters.

5. The system as recited in claim 4, wherein said processing means further includes means for identifying said unknown character based on the surrounding context of said medium.

6. The system as recited in claim 5, wherein said system further comprises a third memory means for storing a recognized character.

7. The system as recited in claim 6, wherein said system is further comprised of a bus means for coupling said first memory means, said second memory means, said third memory means and said processing means.

8. The system as recited in claim 3, wherein said means for parsing said bit-mapped image of said medium and providing as output a bit-mapped representation of said medium further includes means for distinguishing material consisting essentially of text or as comprising non-text.

9. A method for performing optical character recognition of textual material on a medium, said medium having at least one array element, said method identifying said array element as consisting essentially of text or as comprising non-text, said identification of said array element as consisting essentially of text, comprising the steps of:

(a) determining whether the number of bits in a first state in a predetermined set of bits of said array element is within a first range of values; and

(b) determining whether the number of phase changes from said first state to a second state of said predetermined set of bits of said array element is within a second range of values.

10. The method as recited by claim 9, wherein said predetermined set of bits is a set of 32 bits, said first range of values is between 2 and 12 and said second range of values is between 8 and 12.

11. In an optical character recognition system, a method for identifying a bit-mapped image of an unknown character, comprising the steps of:

(a) generating polygon representations descriptive of the shape of said unknown character;

(b) generating statistical information from said polygon representations descriptive of the shape of said unknown character, and

(c) analyzing said statistical information to identify said unknown character.

13. The method as recited by claim 11, wherein said step of analyzing said statistical information to identify said unknown character comprises analyzing said statistical information with a plurality of separate processes, each of said plurality of separate processes for identifying a separate shape.

12. In an optical character recognition system, a method for identifying a bit-mapped image of an unknown character, comprising the steps of:

(a) comparing said bit-mapped image of an unknown character with each of a plurality of templates in a template cache;

(b) determining if said bit-mapped image of an unknown character matches any of said plurality of templates in said template cache;

if said bit-mapped image of an unknown character does not match any of said plurality of templates in said template cache;

(c) generating polygon representations descriptive of the shape of said unknown character,

(d) generating statistical information from said polygon representations descriptive of the shape of said unknown character; and

(e) analyzing said statistical information to identify said unknown character.

14. The method as recited by claim 12, wherein said step of analyzing said statistical information to identify said unknown character comprises analyzing said statistical information with a plurality of separate processes, each of said plurality of separate processes for identifying a separate shape.

15. A system for optically scanning a medium for determining if said medium is comprised essential of text of non-text, said medium having thereon an unknown character, said system comprising:

scanning means for scanning said medium, said scanning means providing as output a bit-mapped image of said medium;

a first memory means coupled with said scanning means for storing said bit-mapped image;

processing means for discriminating text from non-text coupled with said first memory means, said processing means including means for generating at least one array element from said bit mapped image, means for determining that the number of on-bits in a first predetermined number of bytes of said array element is greater than or equal to a second predetermined number and less than or equal to a third predetermined number and means for determining that the number of phase changes from on-bits to off-bits in said first predetermined number of bytes of said array element is greater than or equal to a fourth predetermined number and less than or equal to a fifth predetermined number.