Courtesy amount read and transaction balancing system

• US 5,193,121 A
• Filed: 08/15/1991
• Issued: 03/09/1993
• Est. Priority Date: 05/31/1988
• Status: Expired due to Fees

• Current Assignee: Computer Sciences Corporation (DXC Technology Company)
• Original Assignee: TRW Financial Systems Incorporated (Northrop Grumman Corporation)
• Inventors: Johnson, Milton, Tumenbatur, Ihsan, Keret, Ilan, Elischer, Victor P., Aragon, David B.
• Primary Examiner(s): Razavi, Michael T.
• Assistant Examiner(s): COUSO, YON JUNG
• Application Number:
  US07/740,165
• Time in Patent Office: 572 Days
• Field of Search: 382/7, 382/9, 382/18, 382/57, 382/61, 235/454
• US Class Current: 382/138
• CPC Class Codes: G06Q 20/042 characterized in that the payment protocol involves at least one cheque
• CPC Class Codes: G06Q 40/02 Banking, e.g. interest calculation or account maintenance credit or loans G06Q40/03
• 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 30/18 Extraction of features or characteristics of the image

Abstract

A method and an apparatus for processing the predetermined number fields containing the handwritten numeric entries. The numeric entries are subjected to character recognition analysis, and a confidence level is associated with each such numeric field reflecting the degree of confidence with which the apparatus has recognized the numeric dollar amounts. The transaction is automatically balanced from the data in the system by computing a combined value of the recognized dollar amounts making up a transaction and comparing the combined value with the transaction total. This computation and comparison is performed automatically, without the prior intervention or assistance of a human operator to review, verify, correct or otherwise modify the dollar amounts recognized by the system.

Claims

1. A method of processing documents to record handwritten numerical entries, a document to be processed including a plurality of first predetermined fields containing handwritten numerical entries and a second predetermined field containing a numerical entry representing the combined value of the numerical entries of said first predetermined fields, said method comprising the steps of:

locating said predetermined fields;

segmenting said predetermined fields into segments, each containing an individual character of the numerical entry of the associated predetermined field;

subjecting each said character segment to character recognition analysis so as to associate a character therewith;

assigning a character confidence level to each said character representing the accuracy with which the associated character was recognized.

deriving an overall confidence level for each numerical field from the associated character confidence levels;

automatically computing the combined value of the recognized numerical entries for said first predetermined fields;

automatically recognizing the numerical entry value of said second predetermined field;

comparing said computed combined value with the recognized numerical value of said second predetermined field to verify whether said computed combined value and said recognized value match;

if said computed and recognized values fail to match, displaying the images of the numerical entries in the order of increasing overall confidence levels, whereby a key operator may enter numerical entries from the images;

repeating said comparing step after each numerical entry from the images; and

terminating said displaying step when a match is found.

2. The method of claim 1, wherein the images of the numerical entries are displayed one after the other in the order of increasing overall confidence levels.

3. The method of claim 2, wherein said predetermined fields are represented by a pixel array and said segmenting step comprises the steps of:

locating groups of connected black pixels within a predetermined field; and

eliminating selected groups of connected black pixels as insignificant according to predetermined criteria, whereby the remaining groups each comprise a distinct segment.

4. The method of claim 3, wherein said character recognition step comprises the step of subjecting each said remaining group to feature analysis, said analysis comprising the step of associating a characteristic feature vector with each said remaining group.

5. The method of claim 4, wherein the step of assigning a character confidence level comprises the steps of:

providing a dictionary of feature vectors, each feature vector having a known character associated therewith;

defining an assembly of subsets of feature vectors in said dictionary, each said subset having a predetermined confidence level associated therewith; and

,

associating a subset of said assembly with said characteristics feature vector and thereby associating said predetermined subset confidence level with said characteristic feature vector.

6. Apparatus for use with transaction balancing and reconciliation apparatus for processing transaction documents bearing numerical entries, said documents including a plurality of first predetermined fields containing handwritten numerical entries and a second predetermined field containing a numerical entry representing the combined value of the numerical entries of said first predetermined fields, said apparatus comprising:

memory means for storing images of said documents;

field-locating means receiving said document images for locating said predetermined fields in said images;

character segment generating means responsive to said field-locating means for partitioning located predetermined fields into segments, each said segment containing an individual character of said numerical entry;

recognition means receiving said segments for recognizing the individual characters contained therein;

character confidence level means communicating with said recognition means for assigning a confidence level to each said segment indicating the confidence with which the character contained therein is recognized;

field assembly means for assembling individual recognized characters into recognized numerical entries, said field assembly means being adapted to communicate said recognized numerical entries to said transaction balancing and reconciliation apparatus;

field confidence level means responsive to said character confidence levels for assigning overall confidence levels to said recognized numerical entries; and

image sequencing means responsive to said transaction balancing and reconciliation apparatus for presenting the images of said predetermined fields in order of increasing overall confidence level to a display monitor when the associated transaction fails to balance;

whereby an operator may present the numerical entries from said display monitor to said transaction balancing and reconciliation apparatus one by one in order of increasing overall confidence level until the associated transaction balances.