An algorithm for online strokes verification of Chinese characters using discrete features

Abstract

Existing Chinese educational software tools teach the students at the primary or kindergarten levels the stroke movements and stroke order of each Chinese character by animating the individual strokes of the whole character, one stroke at a time. In this paper the author presents an algorithm which allows the student to hand-write a specified character online through the computer, and which checks if the individual stroke movements and the strokes' order are correct. The algorithm proposed here uses low-integer-valued coding to represent two categories of features: primitive-stroke features and character features. The proposed algorithm is capable of identifying the typical errors as incorrect movement for the primitive stroke, incorrect stroke type, incorrect relative lengths, incorrect position of a stroke from the rest, incorrect character which looks similar in appearance, incorrect order of strokes and insufficient or extra strokes.