Syntactic-Semantics 방법에 신경망을 결합한 온라인 한글 필기체 인식 = Recognizing On-Line Handwritten Hangul by using Syntactic-Semantics Method combined with Neural Networks

다국어 초록 (Multilingual Abstract)

This paper presents a cursive on-line handwritten Hangul recognition model which permits cursive writing between graphemes. Add semantics to the basic stroke types possessed by graphemes and expand stroke types. Also, by the design of attribute grammer more accurate feature values will be elicited. The combination of the feature values with genetic algorithm and BP will improve the neural networks model, and this type of learning will increase the recognition rate and reduce the learning time. In this paper, this new on-line handwritten Hangul recognition model is proposed.