A Novel Pre-segmentation Algorithm of Handwritten Chinese Characters in e-Commerce

Abstract

With the development of Chinese E-commerce, handwritten Chinese characters recognition has became an important problem in electronic payment systems and so on. However, the Chinese segmentation is the basis for Chinese character recognition. In the existing Chinese characters segmentation methods, the first step is the presegmentation for Chinese characters image, the second step is further fine segmentation for results of presegmentation, which is a commonly used strategy. Therefore, the pre-segmentation is very important. In order to deal with inefficiency of the traditional algorithms, the paper presented a novel algorithm with pre-segmentation method. Performance analysis and simulation results show that the novel has better efficiency, higher precision and less system cost.