ON-LINE HAND-WRITTEN CHARACTER RECOGNITION METHOD AND CHARACTER DESCRIPTION METHOD FOR ON-LINE HAND-WRITTEN CHARACTER RECOGNITION

摘要

PROBLEM TO BE SOLVED: To reduce the size of a dictionary for on-line hand-written character recognition. ;SOLUTION: The stroke of an input character is normalized by the size of its bounding box, and then the quasi-phase structure is analyzed, and the stoke type of the stroke of the input character is determined by matching between this quasi-phase structure and those of fundamental strokes in a stroke dictionary 114. Stroke types of respective strokes and coordinates of diagonal points or start and end posits of its bounding box are used to identify the character category by matching with character models in a character dictionary 16. Character models are described by stroke types of respective strokes, coordinates of diagonal points or start and end points of bounding boxes, numbers of strokes, and character codes.;COPYRIGHT: (C)1997,JPO