摘要
在对掌纹原始图像进行去噪、分割等预处理之后,利用平移不变的Zernike矩特征矢量(TIZMs)作为掌纹特征建立特征库,根据已知分类信息建立样本集.并将问题分解为多个小规模的两类问题,然后采用模块化神经网络(MNN)作为分类器进行掌纹识别.对香港理工大学的Polyu PalmprintDB数据库中的3200个掌纹进行实验,在响应时间和识别精度等方面获得了很好的结果.
This paper introduces a new approach for palmprint recognition, using translation invariant Zernike moments （TIZMs） as palm features, and a modular neural network （MNN） as classifier. Translation invariance is added to the general Zernike moments which have a good property of rotation invariance. The pattern set is set up by eight-order TIZMs with 25 dimensions. A modular neural network is presented in order to decompose the palmprint recognition task into a series of smaller and simpler two-class sub-problems. Simulations have been done on the Polyu _ PalmprintDB database, which is composed of 3200 palmprints （ 10 palmprints/person）. Experimental results demonstrate that higher identhqcation rate and recognition rate are achieved by the proposed method in contrast with the straight-line segments （SLS） based method and the Fuzzy Directional Element Energy Feature （FDEEF） method.
出处
《高技术通讯》
CAS
CSCD
北大核心
2005年第12期19-23,共5页
Chinese High Technology Letters
基金
中国科学院资助项目，新材料领域项目
