小波包分解在虹膜识别中的应用

Application of wavelet packets decompositions in iris recognition

小波包分解是从小波变换延伸而来的一种更精细的信号分析算法,具有分析高频特征信息的优势。文中结合小波包分解的特点,提出了一种基于小波包分解的虹膜识别算法。该算法首先对虹膜图像实行分窗小波包分解,并对各窗口的子带图像作筛选处理;然后通过奇异值分解对筛选后的各子带图像作进一步的特征提取和压缩,得到虹膜识别特征;最后利用加权欧氏距离分类器进行识别。实验结果表明了该算法的有效性。

The method of wavelet packets decompositions originating from wavelet transform is more accurate in signal analysis, with the predominance of analyzing details information of high frequency. Combined with the traits of wavelet packets decompositions, an algorithm for iris recognition was presented in this paper. Firstly, iris image was divided into several windows, and wavelet packets decompositions were done to them. At the same time, some subbands from each window were selected, which contain most information of iris image. Secondly, the farther feature extraction and compression were applied to these subband images by way of SVD（ Singular Value Decomposition）, and iris recognition features were obtained. Finally, weighted Euclidean distance classifier was utilized in recognition. Experimental results show the method is valid in iris recognition.