一种基于正交不变集多小波的虹膜识别方法

Iris Recognition Algorithm Based on Orthogonal Invariant Set Multi-wavelet

提出了一种新的虹膜特征提取方法。该方法应用双正交不变集多小波对虹膜图像进行特征提取,同时考虑了虹膜图像的局部特征和全局特征,采用相关系数测度和方差倒数加权欧氏距离相结合的方法来设计分类器,为了消除虹膜图像在采集时可能产生光斑而对识别结果带来影响,在设计分类器时采用分块处理的办法。实验结果表明,多小波变换能更好的提取虹膜的纹理特征,获得更好的识别性能,是一种有效识别方法。

A new method for feature extraction of iris image was proposed in this paper.The orthogonal multi-wavelet filter based on the triangle domain was used to extract the features from the iris images.Global and local texture features of iris were considered when the classifier was designed by combining the correlation coefficient measure and Euclidean distance of covariance reciprocal with weight value.In order to eliminate the effect of facula when the iris images were gathered,the images of iris was divided into small different pieces when the classifier was designed.The experimental results show that multi-wavelet is good at extracting feature of iris,and can achieve better performance than scalar wavelet.The proposed algorithm is valid in iris recognition.