Value based scoring for optical character recognition

Abstract

A technique is described for producing field recognition scores that vary according to the position of the errors. In particular, a value-based score for a recognition result is determined from both the individual character recognition scores and an “error value” that is associated with each character position of the recognition result. In general, this value-based score is computed by adding the error value for each character position of the recognition result. As illustration, an OCR system for scanning checks is adapted to produce field recognition scores that reflect economic value. In this approach, an error in the leftmost digit is “worth” much more than an error in the rightmost digit.

Dependencies are inferred from the claim text.