Address block location using character recognition and addresssyntax




The authors shown many advantages of using postal syntacticfeatures in address block location. Performing character recognition atlow resolution (100 PPI) is desirable for a real-time system. Anefficient word matching algorithm is designed to match words against adictionary of about 500 keywords and assign a syntactic tag to each wordin a text block. The critical information and the street syntax areextracted by parsing the syntactic tags. The usage of the postalsyntactic features has been introduced in many aspects such asrecovering lost information, selecting a more likely destination addressblock, and improving destination address block segmentation. Withcharacter recognition available in the system, the utilization of othersyntax and context information on the mail pieces becomes feasible. Theinitial testing results of the incorporation of the syntactic processingin address block location were positive and encouraging
机译:作者展示了使用邮寄句法的许多优点 地址块位置中的功能。表演字符识别 实时系统期望低分辨率(100ppi)。一个 有效的单词匹配算法旨在匹配对A的单词 大约500个关键字的字典并将语法标记分配给每个单词 在文本块中。关键信息和街道语法是 通过解析语法标记来提取。邮政的使用 在许多方面都引入了句法特征 恢复丢失的信息,选择更可能的目的地地址 块,提高目标地址块分割。和 在系统中可用的字符识别,利用其他 邮件部分的语法和上下文信息变得可行。这 术语概念的句法处理的初步测试结果 在地址块位置是积极和令人鼓舞的



  • 外文文献
  • 中文文献
  • 专利


京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号