Post-processing error correction scheme using a dictionary for on-line handwriting recognition

• US 5,787,197 A
• Filed: 03/28/1994
• Issued: 07/28/1998
• Est. Priority Date: 04/09/1992
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Beigi, Homayoon Sadr Mohammad, Wenstrup, Kenneth Steven, Fujisaki, Tetsunosuke, Modlin, William David
• Primary Examiner(s): Boudreau, Leo H.
• Assistant Examiner(s): MEHTA, BHAVESH M
• Application Number:
  US08/218,996
• Time in Patent Office: 1,583 Days
• Field of Search: 382/186-189, 382/309-311, 382/224-231, 178/18-19, 345/173, 345/179
• US Class Current: 382/187
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/268 Lexical context

Abstract

A dictionary based post-processing technique for an on-line handwriting recognition system is described. An input word has all punctuation removed, and the word is checked against a word processing dictionary. If any word matches against the dictionary, it is verified as a valid word. If it does not verify, a stroke match function and a spell-aid dictionary are used to construct a list of possible words. In some cases, the list is appended with possible words based on changing the first character of the originally recognized word. A character-match score, a substitution score and a word length are assigned to the items on the list. A word hypothesis is constructed from the list with each such word being assigned a score. The word with the best score is chosen as the output word for the processor.

Claims

1. A method of using a dictionary for on-line handwriting recognition, said method comprising the steps of:

inputting a handwritten word to be recognized;

providing a top answer word for recognition, where said top answer word is made up of a sequence of at least one character which is made up of a sequence of at least one segment;

providing a cache of ranked strokes derived from recognition of the handwritten word;

determining if the sequence of characters in the top answer word matches a word in the dictionary with the same spelling, and if so, providing the top answer word as an output word; and

if not

calculating a shape match score for the characters of the top answer word;

determining if the shape match score for the worst character in the top answer word is better than a predetermined threshold, and if so, providing the top answer word as an output word; and

if not

identifying a set of new candidate words from the dictionary, said set comprising words formed from all combinations of segments in the cache that produce recognizable characters;

calculating match scores for the new candidate words, each match score comprising the sum of shape match scores of characters in the new candidate words;

comparing the new match scores of the new candidate words and the top answer and providing the one having the best score as an output word.