Resolution of case confusions by majority voting rule in on-line handwriting recognition.

摘要

In an on-line handwriting recognition system, case confusion is resolved by a majority voting rule. A postprocessing case model makes alphabetic/numeric and upper case/lower case decisions relative to each character in a given word. A word received by a postprocessor counts the non-case-confusable members for each character in a given class. According to the majority of non-case-confusable occurrences for each class, a decision is made whether most of the characters in the word are numeric digits or alphabetic letters, and if letters whether mostly upper case or lower case. The word is then processed again to determine if it should be converted to a similar prototype from a majority case group or whether it should be left unchanged based on the majority of nonconfusable case occurrences. <IMAGE>