Gabor滤波器的掌纹特征提取研究  

Study of Gabor filters on palmprint feature extraction

作　　者：梅支礼 陶海军[1] 王加强[2] 

机构地区：[1]中国计量学院信息工程学院,浙江杭州310018 [2]北京市新技术应用研究所,北京100094

出　　处：《中国计量学院学报》2015年第3期347-352,共6页Journal of China Jiliang University

基　　金：北京市科技计划项目(No.Z131100004113001)

摘　　要：掌纹识别是近年发展起来的一种生物识别技术,掌纹特征提取是掌纹识别核心部分之一,传统Gabor滤波器特征提取速度慢致使掌纹识别速度不能满足实时性的要求.将传统Gabor、改进Gabor和迭代Gabor应用于掌纹特征提取,对比研究了它们在掌纹识别中特征提取速度和识别精度的问题,旨在探求识别精度高且特征提取速度更快的Gabor滤波器,进而改良掌纹特征提取方法.实验表明,迭代Gabor滤波器能确保掌纹识别系统在高识别精度的前提下具有更快的识别速度.Palmprint recognition is a kind of biological recognition technology developed in recent years. Palmprint feature extraction is the core of palmprint recognition. Traditional Gabor filters can not meet the requirement of the real-time palmprint recognition system because of the slow speed of featue extraction, A traditional Gabor, am improved Gabor and a recursive Gabor were used in the palmprint feature extraction in this paper. We studied their speed of feature extraction and the recognition accuracy of palmprint recognition comparatively to find a filter with high recognition accuracy and faster feature extraction speed. The experiments show that the recursive Gabor has the fastest feature extraction speed in the premise of ensuring palmprint recognition accuracy under high recognition accuracy.

关 键 词：掌纹识别 特征提取 迭代Gabor 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]