In order to improve Chinese text recognition rate, in this paper we present a post processing method of corpus-based statistical probabilities. The method has used contextual information more than the lexical lever knowledge.
为了改善汉字文本识别率,本文提出了一种基于语料库统计概率的后处理方法,该方法利用上下文相关信息,超过词汇。
参考来源 - 利用上下文相关信息的汉字文本识别 in C·2,447,543篇论文数据,部分数据来源于NoteExpress
自然背景中的文本识别具有巨大的应用价值。
Text recognition in natural scenes is very helpful to many important applications.
文本识别流的自动校对成为了亟需解决的问题。
So the automatic proofreading has being an urgent problem to solve.
用于将文本识别为地址的智能标记操作,包括地图和驱车方向查询。
Smart tag actions for text recognized as addresses, including map and driving direction lookup.
应用推荐