RECOGNITION OF HANDPRINTED CHINESE CHARACTERS VIA STROKE RELAXATION

A new relaxation matching method based on the information of the neigh borhood relationship among extracted sub-strokes is proposed to recogn ize handprinted Chinese characters (HCCs). In order to ensure the conv ergence in the relaxation process, a new iterated scheme is devised. A supporting function is also designed to solve the problem of wide var iability among writers and some inevitable defects in the preprocessin g procedure. The distance function on which the matching possibilities of sub-strokes are reflected is determined by using the linear progra mming method to obtain the best result. The experiments are conducted by using the Kanji of the ETL-8 database. From the experimental result s, it is shown that the proposed algorithm does improve the recognitio n rate of HCCs.