联机手写维文字符切分算法研究

随着手机短信业务普及,智能手机中实现维吾尔文输入、输出已经是新疆地区1000多万少数民族用户迫切的需求。在连续输入的维吾尔文文章或单词中,切分出一个个的字母,供后续的字母识别使用,字母切分是手写输入识别的核心关键技术。手写维文字符串的分割与字符识别密切相关。采用基于识别的分割方法,系统先通过粗略的图像分析寻找所有可能的切点,在分割的过程中引入识别机制来识别分割碎片,将识别结果经过差值运算后置为每个识别对象的识别可信度,利用移动窗口法找到最佳分割路径。在分类器训练时,采用特征提取来估计分类器参数,得到了性质良好的分类器,试验表明,字符切割准确率高达97.3%。

Method for On_line handwritten Uyghur string segmentation

With the development of SMS,Uyghur input and output are the the need of more than ten millions in the region of ethnic minority by intelligent mobile At consecutive input articles or words,cutting a separate character is used for following letters recognition.Character segmentation are the core of handwriting recognition key technologies.The segmentation of handwritten Uyghur string is closely linked with character recognition.The recognition-based method is used to solve the segmentation problem of on_line...