System and method for correction of optical character recognition with display of image segments according to character data

• US 5,455,875 A
• Filed: 08/03/1993
• Issued: 10/03/1995
• Est. Priority Date: 12/15/1992
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Kolsky, Amir, Medan, Yoav, Walach, Eugene, Heilper, Andre, Chevion, Dan, Kagan, Oren, Gilat, Ittai
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): Johns, Andrew W.
• Application Number:
  US08/100,941
• Time in Patent Office: 791 Days
• Field of Search: 382/57, 382/14, 382/9, 382/36, 382/39
• US Class Current: 382/311
• CPC Class Codes: G06V 10/987 with the intervention of an operator

Abstract

A data entry system generates an electronically stored coded representation of a character sequence from one or more electronically stored document images. The system comprising optical character recognition logic for generating, from the document image or images, character data specifying one of a plurality of possible character values for corresponding segments of the document images. The system also has an interactive display means for generating and sequentially displaying, one or more types of composite image, each composite image comprising segments of the document image or images arranged according to the character data, and a correction mechanism responsive to a user input operation to enable the operator to correct the character data associated with displayed segments.

Claims

1. A data entry system for generating an electronically stored coded representation of a character sequence from one or more electronically stored document images, comprising:

an optical character recognition logic means for generating, from a document image and storing in a data base, character data specifying one of a plurality of possible character values for corresponding segments of the document image;

the optical character recognition means generating a high or low confidence factor for each character;

preprocessing means for generating composite images of possible character values for corresponding segments of a plurality of document images and pointers to characters in the data base for updating such character data;

the preprocessing means arranging the segments in one type of the composite images which consists of groups of segments for which the character data specifies the same character value and has a high confidence factor;

an interactive display means for generating and sequentially displaying, one or more types of composite images, each composite image comprising segments of the document image arranged according to like character data;

the interactive display means having a correction mechanism responsive to a user input operation to enable the operator to correct the character data associated with the displayed segments; and

post processing means for updating character data in the data base corrected by the operator using the pointer for the character data in the data base.

2. The data entry system in claim 1 wherein each character data has a confidence value associated with each specified character value, where the confidence value is indicative of the likelihood that the specified character value is correct, wherein the composite images comprise segments having corresponding confidence value indicating a low such likelihood.

3. The data entry system in claim 2 wherein the segments are arranged in groups of n segments, wherein the correction mechanism enables the operator to set the confidence value of each of the n segments in a group to indicate a high likelihood that the character code is correct through a single user input operation.

4. The data entry system in claim 3 wherein the character data are arranged in fields each comprising one or more characters, the character data comprising a type value associated with each field, and wherein one type of the composite images comprise groups of segments associated with fields having the same type value associated therewith.

5. The data entry system in claim 4 wherein the predetermined user input operation is actuation of a mouse at a time when a screen cursor is located over the displayed segment.

6. The Data entry system in claim 5 wherein the display of the composite images is via a video display unit.

7. A method of operating a data entry apparatus to generate an electronically stored coded representation of a character sequence from an electronically stored image of a document, the method comprising the steps:

generating in optical character means character data specifying one of a plurality of possible character values for corresponding segments of the document images using optical character recognition logic;

the character data comprising a confidence value for each specified character value, the confidence value being indicative of the likelihood that the character code associated therewith is correct, the method further comprised in the steps;

storing the character data and generating a pointer to the character data in a data base;

generating and sequentially displaying, one or more types of composite images from the character values and the document image in a preprocessing module, each composite image comprising the segments of the document image arranged according to the character data;

generating and sequentially displaying a plurality of composite images of a first or exemption data entry type wherein the arrangement of the segments is into groups of segments for which the character data specifies the same character value;

selecting segments displayed in the first type of composite image and setting the corresponding confidence value to indicate a high likelihood of the specified character value being correct;

generating and sequentially displaying a plurality of composite images of a second or memory span data entry type comprising segments having a confidence value indicating a low likelihood associated therewith;

selecting segments of the document image displayed as part of the composite images of the second type;

correcting the character data corresponding to the selected segments to generate the coded representation; and

updating the character data in the data base using the corrected character data and the pointer to the character data.

8. The method of claim 7 in which the character data is arranged in fields each comprising a one or more characters and comprises a type value associated with each field, which type value can take one of a plurality of values, the method further comprising:

generating and sequentially displaying a plurality of composite images of a third or field data entry type comprising groups of segments associated with fields having the same type value associated therewith;

selecting segments of the document image displayed as part of the composite images of the third type; and

correcting the character data corresponding to the selected segments to generate the coded representation.