Batched character image processing

• US 5,696,854 A
• Filed: 06/07/1995
• Issued: 12/09/1997
• Est. Priority Date: 07/26/1993
• Status: Expired due to Fees

• Current Assignee: David H Shepard
• Original Assignee: Cognitronics Imaging Systems Inc.
• Inventors: Shepard, David H.
• Primary Examiner(s): Mancuso, Joseph
• Assistant Examiner(s): DEL ROSSO, GERARD DOMNICK
• Application Number:
  US08/477,921
• Time in Patent Office: 916 Days
• Field of Search: 382/309-311, 382/173, 382/177-179, 382/100, 382/229, 382/317
• US Class Current: 382/311
• CPC Class Codes: G06V 10/987 with the intervention of an operator

Abstract

Character recognition processing wherein each of a batch of documents is scanned to produce corresponding scan data signals forming a rectilinear data array of binary bits at the intersections of a rectangular coordinate grid. These signals are stored and processed by a recognition algorithm to produce identity signals for recognized characters. Groups of non-recognized characters are presented simultaneously to permit rapid identification by inspection. The identification of recognized characters is verified at high speed by simultaneously presenting the character images as respective groups sorted to have the same recognized identities.

Claims

1. For use in a process for recognizing characters on a batch of documents wherein at least a portion of a document is scanned to produce and store scan data signals reflecting the presence of character elements at particular positions on the document;

the characters represented by said stored scan data signals being initially analyzed by recognition procedures including that of processing the stored character scan data signals by a recognition algorithm to establish initial identifications of the characters;

the method of verifying the identity of characters initially identified by said procedures comprising the following steps;

(1) storing character images with purposely inserted bogus identities to be presented together with initially recognized characters to develop verification accuracy statistics;

(2) sorting the stored initially recognized characters and characters with bogus identities into groups wherein the characters of each group have a common identity characterization;

(3) simultaneously presenting images of a number of characters of at least one of said groups;

(4) determining by inspection the presence of a character in at least one of said presented groups failing to have the identity characterization common to the group to which said character belongs;

(5) determining by inspection the correct identity of a character failing to have the identity characterization common to the group to which said character belongs;

(6) developing a corrected identity signal for such correct identity; and

(7) utilizing said corrected identity signal for an initially-incorrectly-identified character to create at least part of an output text for said batch of documents.

2. For use in a process for recognizing characters on documents wherein at least a portion of a document is scanned to produce and store scan data signals reflecting the presence of character elements at particular positions on the document and presenting the images of the scanned characters on the document;

the characters represented by said stored scan data signals being initially analyzed by recognition procedures including that of processing the stored character scan data signals by a recognition algorithm;

the method of verifying the identity of characters initially recognized by said procedures comprising the following steps;

(1) storing with said scan data signals additional data signals representing images of characters with deliberately incorrect identities;

(2) sorting the initially recognized and said additional image data signals into groups with common identity characterizations;

(3) simultaneously displaying images of characters of at least one of said sorted groups;

(4) determining by inspection the presence of characters in one of said displayed groups failing to have the identity characterization common to said one displayed group;

(5) determining by inspection the correct identity of displayed characters failing to have the identity characterization common to said one displayed group; and

(6) developing a corrected identity signal for displayed characters determined in step (5) above.

3. In a method for verifying initially determined identities of characters, the steps of:

storing data representing the images of a number of characters which have been initially identified;

storing data representing the images of characters which are purposely incorrectly identified;

sorting said stored image data into groups of characters of the same identification;

simultaneously displaying a number of character images from a group of the same identification;

finding by inspection the presence of characters in the displayed group failing to have the characterization common to that group; and

determining whether characters which were purposely-incorrectly-identified were found in the preceding step.

4. The method of claim 3, wherein purposely-incorrectly-identified characters are taken from the stored character images the identities of which have been initially determined;

and including the steps of;

altering the identity assigned to each such taken character; and

storing the now incorrectly-identified taken character with other initially-identified character images.

5. The method of claim 3, wherein purposely-incorrectly-identified characters are inserted in a predetermined numerical relationship with respect to the to initially-identified characters.

6. The method of claim 3, including developing verification accuracy statistics based on the results of said step of determining whether purposely-incorrectly-identified characters were found.