关于汉字识别粗分类并行算法的研究

Parallel Algorithms For Prelimirary Classification In Chinese Character Recognition

利用汉字外围轮廓特征进行粗分类是汉字识别中有效的粗分类技术,为了提取一个汉字的外围轮廓特征,本文提出了两种VLsI并行算法和结构:一个是用树网实现的并行算法;另一个是带有树结构的线性脉动阵列,同对分析了该两种算法的时间复杂度。

The prelimirary classification using peripheral features is a useful classification technology in Chinese character recognition. In this paper, for extracting peripheral contour of Chinese character two VLSI parallel algorithms and architectures are presentedt a parallel algorithm on MT and a Systolic algorithm on 1-D array with tree organization.And then, we have also analized their time complexities.