METHOD AND DEVICE FOR ROUGH CLASSIFYING INPUT CHARACTERS FOR ON-LINE CHARACTER RECOGNITION

摘要

PROBLEM TO BE SOLVED: To not limit the number and the order of strokes of an input character by performing a rough classification while complementarily using a feature amount calculated from feature points extracted from the number of strokes in the input character and the coordinate data sequence of the input character. ;SOLUTION: First of all, stroke counting processing 12 calculates the number of strokes from the coordinate data sequence of the input character containing the ON/OFF information of a pen provided by coordinate data sequence generating processing 11. Next, by a normalizing processing 13 the coordinate data sequence of the input character is normalized, and feature point extracting processing 14 extracts the feature points. Afterwards, the feature amount provided from the feature points is calculated by feature amount calculating processing 15, and in a rough classifying processing 16, the provided number of strokes and feature amount are compared with the number of strokes and a feature amount described in a rough classification dictionary 17 so that candidate recognized characters can be selected and rough input character classification can be performed.;COPYRIGHT: (C)1997,JPO