CAJ | 학술논문

针对联机手写字符的特点,有效地将脱机手写字符的结构特征和统计特征与联机手写字符的笔划特征联系起来,使用内部圆个数,宽高比,密度,射线个数,线段,尖角,斜率七种特征形成模板,使用贝叶斯网络进行培训并识别.实践证明了该方法的可行性.
침대련궤수사자부적특점,유효지장탈궤수사자부적결구특정화통계특정여련궤수사자부적필화특정련계기래,사용내부원개수,관고비,밀도,사선개수,선단,첨각,사솔칠충특정형성모판,사용패협사망락진행배훈병식별.실천증명료해방법적가행성.
Due to the characteristic of on-line, the structure and statistical characteristic of handwritten characters off-line are combined effectively with the stroke characteristics of handwritten characters on-line, the template is formed with seven charac-teristics, including the number of internal circles, aspect ratio, density, the number of ray, segment and slope, and Bayesian net-work is used to train and recognize. Practice has proven that the method is feasible.