基于Gabor小波和增强Fisher模型的掌纹特征提取

对于其他身份识别的生物特性而言掌纹有着很多的先天优势,因此在各个领域得到了非常广泛的应用。本文通过研究掌纹的特征,提出了一种基于Gabor小波和增强Fisher线性判别模型(EFM)的掌纹特征提取算法。先对预处理掌纹,在对掌纹灰色图像通过算法进行Gabor小波变换后,得到Gabor的掌纹特征向量。随后,通过主成分分析变换高维特征向量至低维空间,最后在此空间内利用EFM变换矩阵提取到掌纹的特征。由于Gabor函数在特征提取方面有着优良的性能,对高维特征的降维处理问题可有效解决,同时,算法也提高了Fisher线性判别式(FLD)的推广能力,可以较好地实现掌纹的特征提取。

Palm Print Feature Extraction Based on Gabor Wavelet and Enhanced Fisher Discriminant Model

There are many advantages for the biological characteristics of other identification in terms of palmprint,so it has been widely used in various fields.In this paper,the characteristics of palmprint,proposed one kind based on the Gabor wavelet and enhanced Fisher linear discriminant model(EFM) Palmprint Feature Extraction algorithm.The first treatment of the palmprint,on the palm gray image by Gabor wavelet transform based algorithm,get the palmprint feature vector Gabor.Then,through the principal component analysis transform high-dimensional feature vectors into a low dimensional space,then the space by EFM transform matrix to extract the palmprint features.Due to the Gabor function in the aspect of feature extraction with excellent performance,dimensionality reduction of high dimensional feature can be solved effectively,at the same time,the algorithm can improve the Fisher linear discriminant(FLD) generalization ability,can better achieve the Palmprint Feature extraction.