Character recognition, including method and system for processing checks with invalidated MICR lines

• US 6,654,487 B1
• Filed: 03/03/2000
• Issued: 11/25/2003
• Est. Priority Date: 03/03/2000
• Status: Expired due to Term

• Current Assignee: Advanced Financial Solutions Incorporated (Fidelity National Information Services Incorporated)
• Original Assignee: Advanced Financial Solutions Incorporated (Fidelity National Information Services Incorporated)
• Inventors: Downs, Charles H. Jr.
• Primary Examiner(s): Au, Amelia M.
• Assistant Examiner(s): MILLER, MARTIN E
• Application Number:
  US09/518,424
• Time in Patent Office: 1,362 Days
• Field of Search: 382/137, 382/138, 382/139, 382/140, 382/310, 382/318, 382/258, 382/259, 382/209, 382/213
• US Class Current: 382/139
• CPC Class Codes: G06V 10/98 Detection or correction of errors, e.g. by rescanning the pattern or by human intervention; Evaluation of the quality of the acquired patterns

Abstract

An automated analysis method and tool for a digitally imaged financial item can digitally recognize a plurality of characters across an area of a digital image containing character indicia at predetermined positions in the area and also possibly containing non-character indicia in the area. This finds exemplification in an automated check processing method and check sorting system in which a check'"'"'s MICR line that has been read and interpreted as invalid can be digitally analyzed, via a digital image of the check, to identify MICR characters that could not be validly read or interpreted by the conventional MICR processing equipment. This includes digitally applying character recognition processing to an invalidated MICR line in a digital image of the financial item (e.g., a check). Previously invalidated MICR line data can then be corrected.

Claims

1. A check analysis method, comprising:

using a first MICR font template to identify MICR characters across the length of a selected area of the digital image of the check;

detecting a position for a MICR character not identified by the first MICR font template relative to at least one MICR character identified by the first MICR font template; and

in response to detecting a position for a MICR character not identified by the first MICR font template, using another MICR font template to identify the MICR character in the detected position, wherein such other MICR font template includes an inverse skeleton template defining regions outside a MICR character of the first MICR font template within which to identify a MICR character based on pixels outside the MICR character location in the detected position.

2. A check analysis method as defined in claim 1, further comprising changing stored MICR line data for the check in response to identifying sufficient valid MICR characters to comply with predetermined rules defining a valid MICR line for a check.

3. A computer-implemented method of analyzing a MICR line in a digital image, comprising:

scanning the MICR line in the digital image to identify MICR characters therein, wherein scanning the MICR line includes using first MICR font templates to identify MICR characters in the MICR line;

in response to identified characters, determining positions along the MICR line in the digital image where other MICR characters should be; and

analyzing each determined position of the MICR line in the digital image to try to identify the respective MICR character at that position, wherein analyzing each determined position includes using second MICR font templates to try to identify additional MICR characters in the MICR line, and wherein analyzing each determined position further includes using third MICR font templates to try to identify MICR characters in response to failing to identify a MICR character at a respective position using the second MICR font templates, the third MICR font templates including skeleton templates the inverse of at least one of the first MICR font templates and the second MICR font templates for identifying broken MICR characters having a part of the character missing and not identified by either the first or second MICR font templates.

4. A computer-implemented method of analyzing a MICR line in a digital image as defined in claim 3, further comprising digitally removing horizontal and vertical lines detected in the MICR line and not part of a possible MICR character.

5. A check processing method as defined in claim 3, wherein:

using first MICR font templates to identify MICR characters includes identifying acceptable pixels and unacceptable pixels in response to a selected one of the first MICR font templates, subtracting the number of unacceptable pixels from the number of acceptable pixels to define a difference, and comparing the difference to a first predetermined threshold;

using second MICR font templates to try to identify additional MICR characters includes determining a sum of acceptable pixels in response to a selected second MICR font template and comparing the sum of acceptable pixels to a second predetermined threshold; and

using third MICR font templates to try to identify MICR characters includes determining a sum of unacceptable pixels in response to a selected one of the third MICR font templates and comparing the sum of unacceptable pixels to a third predetermined threshold.

6. A check processing method as defined in claim 5, wherein the second predetermined threshold sets a relaxed standard relative to the first predetermined threshold.

7. A computer-implemented method for digitally recognizing a plurality of characters across an area of a digital image containing character indicia at predetermined positions in the area and also containing non-character indicia in the area, comprising:

detecting indicia in the digital image and using a first template on the detected indicia to distinguish at least some character indicia from non-character indicia across the area;

in response to the predetermined positions of the character indicia and the character indicia distinguished using the first template, determining positions in the area at which other character indicia should be located but are not distinguished using the first template; and

using another template at the determined positions to try to identify character indicia thereat, wherein such other template represents an inverse of the first template, the inverse defining an area of investigation outside an area of the first template at the determined positions.

8. A computer-implemented method as defined in claim 7, further comprising digitally distinguishing non-character indicia representing horizontal and vertical lines detected in the area of the digital image.

9. A check analyzer tool, comprising:

a first MICR font template to identify MICR characters;

a missing MICR character position detector responsive to the first MICR font template to detect a position of a missing MICR character relative to at least one MICR character identified by the first MICR font template; and

another MICR font template, for use in response to the missing MICR character position detector to identify a MICR character in apposition detected by the missing MICR character position detector, wherein such other MICR font template represents an inverse of the first MICR font template so that the inverse identifies broken MICR characters having a part of the character missing and not identified by the first MICR font template.

10. A computer-implemented method for digitally recognizing a plurality of characters across an area of a digital image containing character indicia at predetermined positions in the area and also containing non-character indicia in the area, comprising:

detecting indicia in the digital image and using skeleton templates on the detected indicia to distinguish at least some character indicia from non-character indicia across the area;

in response to the predetermined positions of the character indicia and the character indicia distinguished using the skeleton templates, determining positions in the area at which other character indicia should be located but are not distinguished using the skeleton templates; and

using inverse skeleton templates at the determined positions to try to identify character indicia thereat, wherein the inverse skeleton templates represent character indicia inversely to character indicia representations of the skeleton templates such that the inverse skeleton templates define areas of investigation outside areas of the skeleton templates at the determined positions.

11. A financial item analyzer tool, comprising:

means for applying character recognition processing to an invalidated MICR line in a digital image of a financial item; and

memory means for containing in digitally encoded form the means for applying such that the means for applying is accessible by a digital computer;

wherein the means for applying includes;

a first set of digital templates, the first set including skeleton templates representing MICR characters to be digitally recognized at determined positions of the invalidated MICR line; and

a second set of digital templates, the second set including inverse templates representing respective inverses of the MICR characters such that each inverse template of the second set is an inverse of a respective one of the skeleton templates of the first set, wherein a respective inverse template of the second set defines an area of investigation outside an area of the respective skeleton template of the first set at the determined positions.

12. A financial item analyzer tool as defined in claim 11, wherein the financial item is a check passed through a check sorter.

13. A financial item analyzer tool as defined in claim 11, wherein the means for applying includes:

a detector to detect a digital image of the invalidated MICR line in the digital image of the financial item, wherein the detector includes the first and second sets of digital templates; and

an image orienter, responsive to the detector, to change an orientation of the digital image of the financial item.

14. A financial item analyzer tool, comprising:

a first MICR font template including a set of skeleton templates and a first application rule set to identify MICR characters;

a missing MICR character position detector responsive to the first MICR font template to detect a position of a missing MICR character relative to at least one MICR character identified by the first MICR font template;

a second MICR font template, responsive to the missing MICR character position detector, the second MICR font template including the set of skeleton templates and a second application rule set to identify a MICR character in a position detected by the missing MICR character position detector; and

a third MICR font template, responsive to the second MICR font template, the third MICR font template including a set of inverse skeleton templates and a third application rule set to identify a MICR character in a position detected by the missing MICR character position detector.

15. A financial item analyzer tool as defined in claim 14, wherein:

the first application rule set includes a first predetermined threshold, against which a difference between acceptable pixels and unacceptable pixels determined with the first MICR font template is compared;

the second application rule set includes a second predetermined threshold, against which a sum of acceptable pixels determined with the second MICR font template is compared; and

the third application rule set includes a third predetermined threshold, against which a sum of unacceptable pixels determined with the third MICR font template is compared.

16. A financial item analyzer tool as defined in claim 15, wherein the second predetermined threshold sets a relaxed standard relative to the first predetermined threshold.

17. A check processing method, comprising:

transporting a check through a check sorter including a MICR reader and a digital imager;

generating electrical signals with the MICR reader in response to the MICR reader sensing a MICR line on a check in the check sorter and communicating the electrical signals to a controller including a programmed computer to define a first set of data;

generating with the digital imager a digital image of the check in the check sorter and communicating the digital image to a database to define a second set of data;

determining in the programmed computer, using the first set of data and a predetermined set of MICR line validation rules digitally stored in the programmed computer, whether the electrical signals represent a valid or an invalid MICR line; and

in response to determining that the electrical signals represent an invalid MICR line, and off-line from the check sorter and the transporting of the check, digitally processing, using the second set of data, the digital image of the check to identify MICR characters therein, including;

digitally converting a digital representation of the digital image from the second set of data to a common image format;

digitally applying sets of skeleton templates to the converted data in the common image format to determine MICR characters in the MICR line represented in the converted data;

determining in the programmed computer, using the predetermined set of MICR line validation rules, whether the determined MICR characters represent a valid or invalid MICR line; and

writing the determined MICR characters to the first set of data in response to determining that the determined MICR characters represent a valid MICR line;

wherein digitally applying sets of skeleton templates includes;

using first MICR font skeleton templates to identify MICR characters across the length of a selected area of the digital image of the check represented in the converted data in the common image format;

detecting a position for a MICR character not identified by the first MICR font skeleton templates relative to at least one MICR character identified by the first MICR font skeleton templates; and

in response to detecting a position for a MICR character not identified by the first MICR font skeleton templates, using other MICR font templates to identify the MICR character in the detected position, wherein such other MICR font templates are inverses of the first MICR font skeleton templates.

18. A check processing method, comprising:

transporting a check through a check sorter including a MICR reader and a digital imager;

generating electrical signals with the MICR reader in response to the MICR reader sensing a MICR line on a check in the check sorter and communicating the electrical signals to a controller including a programmed computer to define a first set of data;

generating with the digital imager a digital image of the check in the check sorter and communicating the digital image to a database to define a second set of data;

determining in the programmed computer, using the first set of data and a predetermined set of MICR line validation rules digitally stored in the programmed computer, whether the electrical signals represent a valid or an invalid MICR line; and

in response to determining that the electrical signals represent an invalid MICR line, and off-line from the check sorter and the transporting of the check, digitally processing, using the second set of data, the digital image of the check to identify MICR characters therein, including;

digitally converting a digital representation of the digital image from the second set of data to a common image format;

digitally applying sets of skeleton templates to the converted data in the common image format to determine MICR characters in the MICR line represented in the converted data;

determining in the programmed computer, using the predetermined set of MICR line validation rules, whether the determined MICR characters represent a valid or invalid MICR line; and

writing the determined MICR characters to the first set of data in response to determining that the determined MICR characters represent a valid MICR line;

wherein digitally applying sets of skeleton templates includes;

using first MICR font skeleton templates to identify MICR characters in the converted data;

using second MICR font skeleton templates to try to identify additional MICR characters in the MICR line, and

using third MICR font templates to try to identify still other MICR characters, the third MICR font templates including templates that are inverses of the first and second MICR font skeleton templates for identifying broken MICR characters having a part of the character missing and not identified by either the first or second MICR font templates.

19. A check processing method as defined in claim 18, wherein:

using first MICR font skeleton templates to identify MICR characters in the converted data includes identifying acceptable pixels and unacceptable pixels in response to a selected one of the first MICR font skeleton templates, subtracting the number of unacceptable pixels from the number of acceptable pixels to define a difference, and comparing the difference to a first predetermined threshold;

using second MICR font skeleton templates to try to identify additional MICR characters in the MICR line includes determining a sum of acceptable pixels in response to a selected second MICR font skeleton template and comparing the sum of acceptable pixels to a second predetermined threshold; and

using third MICR font templates to try to identify still other MICR characters includes determining a sum of unacceptable pixels in response to a selected one of the third MICR font templates and comparing the sum of unacceptable pixels to a third predetermined threshold.

20. A check processing method as defined in claim 19, wherein the second predetermined threshold sets a relaxed standard relative to the first predetermined threshold.

21. A check processing method, comprising:

transporting a check through a check sorter including a MICR reader and a digital imager;

generating electrical signals with the MICR reader in response to the MICR reader sensing a MICR line on a check in the check sorter and communicating the electrical signals to a controller including a programmed computer to define a first set of data;

generating with the digital imager a digital image of the check in the check sorter and communicating the digital image to a database to define a second set of data;

determining in the programmed computer, using the first set of data and a predetermined set of MICR line validation rules digitally stored in the programmed computer, whether the electrical signals represent a valid or an invalid MICR line; and

in response to determining that the electrical signals represent an invalid MICR line, and off-line from the check sorter and the transporting of the check, digitally processing, using the second set of data, the digital image of the check to identify MICR characters therein, including;

digitally converting a digital representation of the digital image from the second set of data to a common image format;

digitally applying sets of skeleton templates to the converted data in the common image format to determine MICR characters in the MICR line represented in the converted data;

determining in the programmed computer, using the predetermined set of MICR line validation rules, whether the determined MICR characters represent a valid or invalid MICR line; and

writing the determined MICR characters to the first set of data in response to determining that the determined MICR characters represent a valid MICR line;

wherein digitally applying sets of skeleton templates includes;

detecting indicia in the converted data and using a first template on the detected indicia to distinguish at least some character indicia; and

using another template on detected indicia to try to identify character indicia, wherein such other template represents an inverse of the first template, the inverse defining regions outside a MICR character to be examined.

22. A check processing method, comprising:

transporting a check through a check sorter including a MICR reader and a digital imager;

generating electrical signals with the MICR reader in response to the MICR reader sensing a MICR line on a check in the check sorter and communicating the electrical signals to a controller including a programmed computer to define a first set of data;

generating with the digital imager a digital image of the check in the check sorter and communicating the digital image to a database to define a second set of data;

determining in the programmed computer, using the first set of data and a predetermined set of MICR line validation rules digitally stored in the programmed computer, whether the electrical signals represent a valid or an invalid MICR line; and

in response to determining that the electrical signals represent an invalid MICR line, and off-line from the check sorter and the transporting of the check, digitally processing, using the second set of data, the digital image of the check to identify MICR characters therein, including;

digitally converting a digital representation of the digital image from the second set of data to a common image format;

digitally applying sets of skeleton templates to the converted data in the common image format to determine MICR characters in the MICR line represented in the converted data;

determining in the programmed computer, using the predetermined set of MICR line validation rules, whether the determined MICR characters represent a valid or invalid MICR line; and

writing the determined MICR characters to the first set of data in response to determining that the determined MICR characters represent a valid MICR line;

wherein digitally applying sets of skeleton templates includes digitally using in the programmed computer a digital inverse skeleton template representing an inverse image of the character to be detected such that a broken representation of the character in the digital image is identified.