Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Method for Locating and Correcting Optical Character Recognition Errors

IP.com Disclosure Number: IPCOM000105639D
Original Publication Date: 1993-Aug-01
Included in the Prior Art Database: 2005-Mar-20
Document File: 2 page(s) / 57K

Publishing Venue

IBM

Related People

Geldner, JF: AUTHOR

Abstract

With the increase of image systems in every form of industry, there is a growing requirement for optical character recognition to convert the image data into character data.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Method for Locating and Correcting Optical Character Recognition Errors

      With the increase of image systems in every form of industry,
there is a growing requirement for optical character recognition to
convert the image data into character data.

     Image systems have been more highly integrated into the mission
critical systems of many major industries.  To fully exploit the
advantages of an image system, many businesses have implemented
optical character recognition (OCR) to convert image data into
character data.  The character data can then be used as input to the
businesses' existing applications.

     Due to the imperfect nature of OCR systems there exists a need
for a facility to correct OCR errors to ensure the accuracy of the
character data.  OCR errors can result from one of the following
instances.  First, if the OCR system does not recognize the image
data as a valid character, and second, if the OCR system substitutes
the wrong character for the image data.  Currently, the bottleneck in
processing OCR data is the correction process.  This bottleneck
exists due to the fact that character data can be produced by an OCR
system at a faster rate than the human can correct the character
data.  To achieve the most effective image system, the error
correction process time must be minimized to provide maximum
efficiency.

     This invention will provide a method for increasing the error
correction throughput by minimizing the time necessary to correct OCR
errors.  Edit windows that contain the actual c...