基于统计的动态规划算法在联机汉字切分中的应用  

Statistics based dynamic programming method for online handwritten character segmentation

作　　者：姚正斌[1] 丁晓青[1] 刘长松[1] 

机构地区：[1]清华大学电子工程系智能系统与技术国家重点实验室,北京100084

出　　处：《计算机应用研究》2011年第7期2592-2594,共3页Application Research of Computers

基　　金：国家"973"计划资助项目(2007CB311004)

摘　　要：针对中文联机手写文本识别中需要对不同长度的切分路径进行快速合理评价的问题,提出了一种基于后验概率的动态规划代价函数计算方法,结合切分块与切分点的几何信息,利用原始切分块的个数对代价值进行加权。实验证明,该方法有效克服了不同路径长度的负面影响,提高了最优切分路径搜索的准确率。To solve the problem that rapid and effective evaluation of the segmentation path is needed for online handwritten Chinese text recognition,this paper proposed a statistics based cost function for dynamic programming.It utilized geometrical information of segmentation blocks and points and weighted by the number of segmentation blocks in the original segmentation path.Experiment proves that this method can overcome the shortcoming caused by the variance of length of segmentation paths and increase the accuracy of optimal path search.

关 键 词：字符切分 动态规划 切分路径 代价函数 后验概率 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]