Abstract: A new iris recognition method searching for the appropriate wavelet scale is proposed using the bandpass characteristic of wavelets and wavelet transform principles for detecting singularities.The iris image texture is divided into eight bands，and 32 scale continuous wavelet transformation coefficients are quantized to iris codes for each analysis band.Then the Hamming distance is adopted to match two iris codes.The appropriate scale for each band is obtained by searching for the minimum from 32 Hamming distances.The results show that this method achieves an identification rate of 98.15%，and its performance is further improved compared with the one previously proposed.It is promising to be used in a large scale personal identification system.

Key words: iris recognition, wavelet transform, feature extraction, Hamming distance

摘要： 利用小波的带通特性和小波变换检测奇异点的原理提取虹膜特征，并在此基础上提出了一种新的搜索合适小波尺度的虹膜识别算法。将虹膜纹理分成8个分析带，对每个分析带进行连续小波变换，取其中32个尺度下的结果量化编码。然后利用Hamming距离进行模式匹配，对每个分析带通过搜索最小的Hamming距离获得合适的尺度。实验结果表明，算法的识别率可达98.15%，同以前提出的算法相比识别性能进一步提高，可用于大规模身份识别系统。

关键词: 虹膜识别, 小波变换, 特征提取, Hamming距离