Orthogonal technology for multi-line character recognition

• US 20020054693A1
• Filed: 07/30/2001
• Published: 05/09/2002
• Est. Priority Date: 07/28/2000
• Status: Active Grant

• Current Assignee: Matthews International Corporation
• Original Assignee: RAF Technology, Incorporated (Constellation Software Incorporated)
• Inventors: Elmenhurst, Brian J.
• Granted Patent: US 7,031,519 B2
• Time in Patent Office: Days
• US Class Current: 382/101
• CPC Class Codes: G06F 18/254 of classification results, e.g. of results related to same input data
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/12 Detection or correction of errors, e.g. by rescanning the pattern
• CPC Class Codes: G06V 30/153 using recognition of characters or words
• CPC Class Codes: G06V 30/1918 Fusion techniques, i.e. combining data from various sources, e.g. sensor fusion

Abstract

The present invention encompasses a self-orthogonal character recognition engine for executing an iterative method employing a database of predetermined character strings. The method receives a digital representation of a character string. It then generates a proposed result string by applying to the captured digital image a predetermined recognition routine including one or more recognition subroutines. Each recognition subroutine employs an initial parameter setting. Next, if the proposed result string does not match any of the predetermined character strings in the database, the initial parameter setting of a recognition subroutine is changed to a next setting. The recognition process is then repeated using the next parameter setting to generate and test a next result string. The process can be repeated iteratively until a result string is verified or the process times out.

Claims

1. An iterative character recognition method employing a database of predetermined character strings, the method comprising the steps of:

(a) receiving a digital representation of a character string;

(b) generating a proposed result string by applying a predetermined recognition routine to the received digital representation, the predetermined recognition routine including a recognition subroutine employing an initial parameter setting;

(c) determining whether the proposed result string matches any of the predetermined character strings;

(d) if the proposed result string does not match any of the predetermined character strings, adjusting the initial parameter setting of the recognition subroutine; and

then

(e) repeating steps (b) and (c) to generate a next proposed result string and to determine whether the next proposed result string matches any of the predetermined character strings.

2. The method of claim 1 further including the step of repeating steps (d) and (e) until reaching a termination condition.

3. The method of claim 2 wherein the termination condition is the proposed result string matching any of the predetermined character strings.

4. The method of claim 2 wherein the termination condition is an expiration of a predetermined duration of time.

5. The method of claim 2 wherein the termination condition is a completion of a predetermined number of repetitions of steps (d) and (e).

6. The method of claim 2 wherein the predetermined recognition routine is a magnetic ink character recognition routine.

7. The method of claim 2 wherein the predetermined recognition routine is an optical character recognition routine.

8. The method of claim 7 wherein the optical character recognition routine is a multi-line optical character recognition routine.

9. The method of claim 7 wherein:

the optical character recognition routine is for mail processing;

the character string is a mailing address; and

the predetermined character strings are known mailing addresses.

10. The method of claim 9 wherein the recognition subroutine is chosen from a group consisting of an address block location subroutine, an image enhancement subroutine, a segmentation subroutine, a feature extraction subroutine, a character selection subroutine, a confidence subroutine, and a handwriting recognition subroutine.

11. The method of claim 10 wherein:

the recognition subroutine is the character selection subroutine;

the initial parameter setting references a first character set; and

the adjusted parameter setting references a second character set.

12. The method of claim 10 wherein:

the recognition subroutine is the address block location subroutine;

the initial parameter setting is a first Boolean value instructing whether to use a predefined address block location assumption; and

the adjusted parameter setting is a second Boolean value instructing whether to use the predefined address block location assumption.

13. The method of claim 10 wherein:

the recognition subroutine is the segmentation subroutine;

the initial parameter setting is a first Boolean value instructing whether to use a heuristic segmentation algorithm; and

the adjusted parameter setting is a second Boolean value instructing whether to use the heuristic segmentation algorithm.

14. The method of claim 1 wherein adjusting the initial parameter setting includes selecting a next parameter setting.

15. The method of claim 1 wherein the step of determining the proposed result string match includes using fuzzy logic.

16. The method of claim 1 wherein the digital representation is a digital image.

19. The character recognition engine of claim 18 wherein the database of acceptable result strings is operated independent of the self-orthogonal character recognition engine.

20. The character recognition engine of claim 18 wherein the database of acceptable result strings is external to the self-orthogonal character recognition engine.

17. A method for defining an iterative character recognition routine for use in a self-orthogonal character recognition engine, the method comprising the steps of:

designating a recognition subroutine for inclusion in the recognition routine;

selecting a first parameter setting and a second parameter setting for the recognition subroutine;

securing access to a database of acceptable result strings; and

implementing control logic whereby the first parameter setting is used to generate a first proposed result string, the first proposed result string is matched against the acceptable result strings in the database, and, if the first proposed result string does not match the acceptable result strings in the database, the second parameter setting is used to generate a second proposed result string.

18. A self-orthogonal character recognition engine comprising:

a recognition subroutine having a plurality of parameter settings for processing an input string;

a configuration file designating an order of use for the plurality of parameter settings;

an interface for validating a proposed result string against a database of acceptable result strings; and

a character recognition routine for selecting a result string by invoking the recognition subroutine to employ the plurality of parameter settings, one at a time according to the designated order of use, to generate a corresponding plurality of proposed result strings;

implementing the interface to validate each of the corresponding plurality of proposed result strings; and

selecting as the result string a validated proposed result string validated by the interface.