Error correction of Korean courtesy amounts in bank slips using rule information and cross-referencing

Abstract

In this paper, we designed a Korean courtesy amounts recognizer which improves error correction results in recognized character strings. We use two properties of Korean character strings for the post-processor of Korean courtesy amounts. First, there are syntactic rules in character strings. Using this property, we can correct errors in Korean courtesy amounts. This kind of error correction is restricted only to the Korean characters representing the unit of the amounts. Second, we can compare the Korean courtesy amount with the legal amount, and vice versa. Using this property, we can correct errors of the Korean characters representing numerals. With the help of these two properties, we designed the rule post-processing algorithm and the cross-referencing algorithm. The first Korean character recognizer shows a 95.49% recognition rate. For a post-processed Korean character string, the recognizer of Korean courtesy amounts shows 99.26% reliability.