【摘要】 本文叙述了一个基于改进的＂汉字识别特征点方法＂的高精度简、繁体印刷体汉字文本识别系统。引入特征点的方向属性，明显地提高了＂汉字识别特征点方法＂的汉字识别率。文中阐述了该系统各主要环节的原理。经过百万汉字真实印刷文本的严格测试，本系统汉字识别率达到９７．８４％。对质量较高的真实印刷文本，汉字识别率达到９９％以上。更多还原

【Abstract】 AbstractThe system presented in this paper is based on an improved version of thefeature-point method. The new version uses the direction properties of feature points and this greatly improves the recognition accuracy of the method. The main parts of the recognition system are discussed in this paper. The system has been tested on millions of characters from common printed materials, and the recognition rate reaches 97.84%. For textwith higher printing quality, the recognition rate is above 99%.更多还原