摘要
用于身份鉴别的掌纹识别为信息安全提供了一种新的方案。提出一种变换域和统计域相结合的掌纹识别方法。对掌纹感兴趣区域(ROI)进行中值滤波再多级小波分解,对所有的高频子图像进行分块,求取每一子块高频系数的均值和方差,它们的组合构成该图像的特征向量,利用简单的最近邻分类器进行分类。运用UST掌纹图像库,对该算法进行了测试。从识别率为95.5%的实验结果看,该方法优于目前在掌纹识别上使用较多的子空间法。
Palmprint recognition for identification provides a new scheme for information security.This paper presents a com-bination method of transform domain and statistic domain for palmprint identification.The method filters Region Of Interest（ROI） of the palmprint with median filtering and decomposes it into several sub-images with the wavelet transform.Then it blocks the high-frequency sub-image.The mean and the variance of high-frequency coefficients for each sub-block are found.Their combination constitutes feature vector for the image.The nearest neighbor classifier is used to classify the images.The method is tested on the basis of UST palmprint image database.From the experimental result of 95.5% recognition rate,the method is better than the sub-space methods that are used for palmprint identification at present.
出处
《计算机工程与应用》
CSCD
北大核心
2011年第24期17-19,22,共4页
Computer Engineering and Applications
基金
国家自然科学基金(No.60972123)
辽宁省教育厅科研项目(No.L2010436)~~
关键词
生物特征识别
掌纹识别
小波变换
分块统计
高频系数
biometrics recognition
palmprint recognition
wavelet transform
block statistic
high-frequency coefficients
