| 基于BP神经网络的手写签名识别方法 |
| |
| 引用本文： | 万莉,卢苇,尹朝庆.基于BP神经网络的手写签名识别方法[J].交通与计算机,2006,24(4):83-85. |
| |
| 作者姓名： | 万莉 卢苇 尹朝庆 |
| |
| 作者单位： | 1. 武汉理工大学,武汉,430063
2. 武汉市公安局,武汉,430022 |
| |
| 摘 要： | 手写字符识别是一个难度很大，但却具有广阔的应用前景的研究课题。提出了一种基于BP神经网络的手写签名识别方法，实验表明，采用一个签名的6个真伪样本训练BP网络后，对该签名真伪的正确识别率达到80％以上；如果使用10个真伪样本训练网络，那么该签名的真伪识别率约达90％。
|
| 关 键 词： | 手写签名 图像处理 特征提取 BP神经网络 |
| 收稿时间： | 2005-08-30 |
| 修稿时间： | 2006-03-10 |
Handwritten Signature Recognition Method Based on BP Network |
| |
| WAN Li,LU Wei,YIN Chaoqing.Handwritten Signature Recognition Method Based on BP Network[J].Computer and Communications,2006,24(4):83-85. |
| |
| Authors: | WAN Li LU Wei YIN Chaoqing |
| |
| Institution: | Wuhan University of Technology, Wuhan 430063, China;Bureau of Public Security of Wuhan Municipality, Wuhan 430022, China |
| |
| Abstract: | Handwritten signature recognition is a research subject, which is very difficult but promising. Through testing, the recognition rate of signature can be more than 80% when six samples of a signature have been trained. When ten samples of a signature have been trained, the recognition rate can reach to 90%. |
| |
| Keywords: | handwritten signature image processing character extraction BP neural network |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
