CAJ | 학술논문

提出联机识别自然手写汉字的FS识别法。在剖析五笔字根结构和编码原则基础上，对五笔字根作适应性改造，将键盘输入技术与联机识别技术有机融合的一种识别体系。在多库识别体系中首次采用层间分级技术。分析和实验表明：充分考虑了自然手写汉字书写习惯和结构特征，系统有较高稳定性。
제출련궤식별자연수사한자적FS식별법。재부석오필자근결구화편마원칙기출상，대오필자근작괄응성개조，장건반수입기술여련궤식별기술유궤융합적일충식별체계。재다고식별체계중수차채용층간분급기술。분석화실험표명：충분고필료자연수사한자서사습관화결구특정，계통유교고은정성。
In this paper, keyboard input techniques are introduced intoonline recognition of Chinese character based on the studying of Five-Strokes Code and a new approach for unconstrained handwritten Chinese character online recognition—FS recognition method is proposed. Besides, the character-root set have to be altered slightly to meet the recognition features.The results of preliminary experiments show that the system has advantages of high degree of steadiness and tolerance to strokes variance.