印刷体维吾尔文识别后处理

摘要

In this paper , we apply the N-gram model and the algorithm of Levenshtein Distance to Printed Uygur character recognition post-processing. The recognition errors of the system of Printed Uygur character recognition is a regular pattern, by setting weigh of the recognition errors in the algorithm of Levenshtein Distance based on the comparison and analysis and class of the recognition errors, the correct rate of the recognition were improved. Finally , the results of the experiments indicate that the method can definitely increase the correct rate of the recognition.%本文主要讨论将N-gram模型与编辑距离算法运用于印刷体维吾尔文识别后处理.由于印刷体维吾尔文识别系统的识别错误有一定规律性,所以研究中对识别错误进行了比较、分析、分类、并在编辑距离算法中加入识别错误的权值,以提高识别的正确率.最后,通过实验证明本算法能有效提高识别的正确率.