Handwritten Chinese Character Segmentation Using Local Potential Threshold and Minimum Potential Search

Human can segment a character from handwritten Chinese character text line having many overlaps or joins. Therefore, a fairly high degree of accuracy in the character segment,ation may be attained by imitating the human visual processing. In the cognitive science field, it is proposed that the method called 'field of induction on the retina' is similar to human subjective image processing. But, there are some problems (overlaps or joins characters) in character segmentation method using field of induction on the retina. To solve these problems, this paper proposes the local potent,ial threshold method to imitate human movement of viewpoint and the minimum potential search method to segment joined characters.