Exploring recognition of off-line handwritten Chinese characters using double ANN (artificial neural network) classifier

被引：0

作者：

Wang, Ge ^{[1
]}

Xie, Songyun ^{[1
]}

Dang, Zheng ^{[1
]}

机构：

[1] Northwestern Polytechnical University, Xi'an 710072, China

来源：

Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University | 2010年 / 28卷 / 04期

关键词：

Character sets - Support vector machines - Character recognition;

D O I：

暂无

中图分类号：

学科分类号：

摘要：

To implement the recognition of off-line handwritten Chinese characters is an arduous task. Most of the existing methods have in mind just the small character set. So we combine a double ANN classifier with double feature extraction to implement the recognition of the bigger character set. Our method extracts the double features of one Chinese character and then uses them as inputs to train the double ANNs in parallel. Finally, the post-processing module chooses a more robust result as the recognition output. The comparison between the recognition rates of our method with those of SVM(support vector machine) method, presented in Tables 1 and 2, shows preliminarily that our method is better than the existing SVM method for the recognition of bigger off-line handwritten Chinese character set.

引用

页码：574 / 578