摘要
利用联机书写的特点，提出了一种以方向码、笔划、笔顺等为提取特征，然后用统计学原理对手写体字符进行分类、识别的方法．通过对0～9九个数字编制各自独立的模块分别进行识别，经测试，识别率达99.8％以上，由于采用统计的方法，识别结果稳定．
This paper presents a new approach for 0n-line handwrittern digitsrecognition which is based on the turning cede and stroke number and stroke sequence fea-ture extraction, and make detendne by statistics, and design model Of itself to recognize,the approach using 0n-line character's drirmation, through testing of sample diglts from0 to 9, the recognition rate have gone beyond the limit 99. 98%, and it is stable for thestatistics being used.
