| Previous | [1] | [2] | [3] | [4] | [5] | [6] | [7] | [8] |
An-Bang Wang, Kuo-Chin Fan and Jun S. Huang*
Institute of Computer Sciences and Information Engineering,
National Central University
Taiwan, Republic of China
*Institute of Information Science, Academia Sinica,
Taiwan, Republic of China
Since Chinese characters are composed of radicals, it is an intuitive to recognize the radicals inside a character before recognizing the whole character. In this paper, a partial matching algorithm, which is based on a modified relaxation method, is proposed to perform radical recognition on scanned input characters. The standard radicals are stored in the database, and thirteen transformations are developed to represent a radical appearing in thirteen different positions (13 types) inside a square character. The matching of a radical is done by modified relaxation and also nuder some constraints on the positions and the number of strokes of a radical appearing as 13 types. Experiments were conducted on 204 constrained handwritten characters, each containing two to five radicals, for matching 16 radical templates. The total recognition rate was 63.3%, but the rate of identifying at least one radical in a character was 97%.
Keywords: handwritten chinese character recognition, radicals, relaxation, partial matching
Received October 6, 1993; revised August 18, 1994.
Communicated by Jhing-Fa Wang.