一种利用灰度信息和二值信息的手写字符分割

Segmentation Approach for Handwritten Characters Based on Gray Scale and Binary Image

在字符识别系统中,字符的有效分割是识别的关键。该文描述了一种两阶段的手写字符分割方法。根据字符倾斜角度将原字符串图像划分出若干个形状不规则的区域,每个区域含有一条分割线;针对现有分割方法的某些局限性,提出采用字符图像灰度信息和二值信息相结合的方式来求取分割区域中的分割路径。实验结果表明了该方法对于手写字符分割的有效性。

In order to avoid the limitation of the existing segmentation algorithms based on gray-scale image or binary image, this paper presents an approach of handwritten characters segmentation which makes use of the characteristics of gray-scale image and binary image. The characters image is divided into some segmentation regions, each of which contains a segmentation path. Viterbi algorithm is employed to detect the non-linear segmentation path in each region for separating the connected characters. The algorithm performs the search process by combining the information of gray scale and binary images. Experimental resuks show the proposed approach is effective for handwritten characters segmentation.