基于神经网络的脱机中文签名鉴别系统的研究  

Off-line Chinese signature recognition system based on neural network

作　　者：王红春[1] 靳斌[1] 樊旭升[1] 马士明[1] 

机构地区：[1]西华大学电气信息学院,成都610039

出　　处：《计算机应用》2008年第9期2389-2391,共3页journal of Computer Applications

摘　　要：将神经网络与签名识别相结合,利用前向多层神经网络的反向传播算法(即BP算法),采用Matlab神经网络工具箱构建用于特征识别的三层前向神经网络,同时使用了基于Gabor和Zernike相结合的特征提取方法,最终识别出待识别的手写签名。实验结果证明识别率可达到93.70%以上,证明了该方法的有效性。This paper tried to combine neural network and signature verification, using the Back Propagation algorithm （BP） of the multi player Neural Network based on matlab neural network box. And a method of feature extraction based on Gabor fliter and Zernike moment was proposed. The experiment shows that the method is more effective and feasible, and recognition rate has been up to 93.70%.

关 键 词：签名鉴别 神经网络 ZEMIKE矩 GABOR滤波器 特征提取 

分 类 号：TP391.43[自动化与计算机技术—计算机应用技术]