Optical character recognition and item matching assisted by progressively decreasing match criteria

Claim 1

A method of matching a plurality of item amounts from individual items stored in a memory associated with a scanner as a first source list with item amounts contained in a second source list, comprising the steps of: a) performing character recognition on scanned item amounts from said individual items to form a first source list and on item amounts scanned from a second source list; b) establishing a plurality of match criteria each allowing for a varying degree of adherence for determining a match between an item amount from said first source list and an item amount contained in said second source list; c) matching said item amounts of said items contained in said first source list with said item amounts contained in said second source list where exact matches occur; d) relaxing at least one of said criteria; e) matching item amounts from said first source list with item amounts contained in sad second source list using the relaxed criteria; f) repeating steps d and e until all of said criteria have been relaxed; g) determining the order of said items on said second source list and correlating said order with the order of item amounts from said first source list; and h) determining a FORWARD or REVERSE match class comprising the steps of: 1) assigning a sequence number to each item amount from said first source list; 2) assigning a sequence number to each item amount on said second source list; 3) determining a sequence delta for each matched set of item amounts; 4) averaging said sequence deltas; 5) calculating the statistical variance of said sequence deltas; 6) inverting said second source list item amounts order; 7) repeating steps 2 through 5; 8) comparing said calculated statistical variances: 9) designating said match class as FORWARD or REVERSE, depending upon which calculation of step 5 resulted in a lower statistical variance result and one of said statistical variance results being at least larger than a predetermined number of times the other statistical variance, thereby establishing matches between item amounts from said first source list and item amounts from said second source list using progressively lower degrees of adherence to said match criteria.

Claim 2

A method of matching a plurality of item amounts from individual items stored in a memory associated with a scanner as a first source list with item amounts contained in a second source list, comprising the steps of: a) performing character recognition on scanned item amounts from said individual items to form a first source list and on item amounts scanned from a second source list; b) establishing a plurality of match criteria each allowing for a varying degree of adherence for determining a match between an item amount from said first source list and an item amount contained in said second source list; c) matching said item amounts of said items contained in said first source list with said item amounts contained in said second source list where exact matches occur; d) relaxing at least one of said criteria; e) matching item amounts from said first source list with item amounts contained in sad second source list using the relaxed criteria; f) repeating steps d and e until all of said criteria have been relaxed; g) determining the order of said items on said second source list and correlating said order with the order of item amounts from said first source list; and h) determining a FORWARD or REVERSE match class comprising the steps of: 1) assigning a sequence number to each item amount from said first source list; 2) assigning a sequence number to each item amount on said second source list; 3) determining a sequence delta for each matched set of item amounts; 4) averaging said sequence deltas; 5) calculating the statistical variance of said sequence deltas; 6) inverting said second source list item amounts order; 7) repeating steps 2 through 5; 8) comparing said calculated statistical variances: 9) designating said match class as FORWARD or REVERSE, depending upon which calculation of step 5 resulted in a lower statistical variance result and one of said statistical variance results being at least larger than a predetermined number of times the other statistical variance, thereby establishing matches between item amounts from said first source list and item amounts from said second source list using progressively lower degrees of adherence to said match criteria.

Claim 3

A method of matching a plurality of item amounts from individual items stored in a memory associated with a scanner as a first source list with item amounts contained in a second source list, comprising the steps of: a) performing character recognition on scanned item amounts from said individual items to form a first source list and on item amounts scanned from a second source list; b) establishing a plurality of match criteria each allowing for a varying degree of adherence for determining a match between an item amount from said first source list and an item amount contained in said second source list; c) matching said item amounts of said items contained in said first source list with said item amounts contained in said second source list where exact matches occur; d) relaxing at least one of said criteria; e) matching item amounts from said first source list with item amounts contained in said second source list using the relaxed criteria; f) repeating steps d and e until all of said criteria have been relaxed; g) determining the order of said items on said second source list; h) correlating said order with the order of items amounts from said first source list; i) further determining a FORWARD or REVERSE match class comprising the substeps of: 1) assigning a sequence number to each item amount from said first source list; 2) assigning a sequence number to each item amount on said second source list; 3) determining a sequence delta for each matched set of item amounts; 4) averaging said sequence deltas; 5) calculating the statistical variance of said sequence deltas; 6) inverting said second source list item amounts order; 7) repeating steps 2 through 5; 8) comparing said calculated statistical variances; 9) designating said match class as FORWARD or REVERSE, depending upon which calculation of step 5 resulted in a lower statistical variance result and one of said statistical variance results being at least larger than a predetermined number of times the other statistical variance, wherein said match criteria includes the following degrees of adherence for determining said match: only item amounts from said first source list with no rejected characters will be considered for matching, and item amounts from said first source list having a rejected character will be considered for matching, thereby establishing matches between item amounts from said first source list and item amounts from said second source list using progressively lower degrees of adherence to said criteria.

Claim 4

A method of matching a plurality of item amounts from individual items stored in a memory associated with a scanner as a first source list with item amounts contained in a second source list, comprising the steps of: a) performing character recognition on scanned item amounts from said individual items to form a first source list and on item amounts scanned from a second source list; b) establishing a plurality of match criteria each allowing for a varying degree of adherence for determining a match between an item amount from said first source list and an item amount contained in said second source list; c) matching said item amounts of said items contained in said first source list with said item amounts contained in said second source list where exact matches occur; d) relaxing at least one of said criteria; e) matching item amounts from said first source list with item amounts contained in said second source list using the relaxed criteria; f) repeating steps d and e until all of said criteria have been relaxed; g) determining the order of said items on said second source list; h) correlating said order with the order of items amounts from said first source list; i) further determining a FORWARD or REVERSE match class comprising the substeps of: 1) assigning a sequence number to each item amount from said first source list; 2) assigning a sequence number to each item amount on said second source list; 3) determining a sequence delta for each matched set of item amounts; 4) averaging said sequence deltas; 5) calculating the statistical variance of said sequence deltas; 6) inverting said second source list item amounts order; 7) repeating steps 2 through 5; 8) comparing said calculated statistical variances; 9) designating said match class as FORWARD or REVERSE, depending upon which calculation of step 5 resulted in a lower statistical variance result and one of said statistical variance results being at least larger than a predetermined number of times the other statistical variance, wherein said match criteria includes the following degrees of adherence for determining said match: only item amounts from said first source list with no rejected characters will be considered for matching, and item amounts from said first source list having a rejected character will be considered for matching, thereby establishing matches between item amounts from said first source list and item amounts from said second source list using progressively lower degrees of adherence to said criteria.