基于Log-Gabor小波相位一致的掌纹线特征提取

Palm-line feature extraction using Log-Gabor wavelet and phase congruency

研究掌纹线的特点,提出了一种基于相位一致的掌纹线特征提取方法,多方向、多尺度的Log-Gabor小波用来计算掌纹图像的相位一致不变量以获得更丰富的掌纹线特征信息和精确的特征定位。方法提取的线特征不仅包含掌纹的结构信息,还具有纹线的强度和宽度信息;此外,由于使用图像的相位信息,方法受图像亮度和对比度影响较小,提取的特征比较稳定。方法在PolyU掌纹库上测试,结果(EER=0.6%)表明线特征提取的良好效果。

We present a novel palm-line extraction approach based on phase congruency. Multborientation and multi-scale Log-Gabor wavelets are used to achieve effective feature information and well feature location accuracy. Palm-line feature extracted by proposed method can describe both structure and strength information of the palmprint. Furthermore, the extracted feature is very stable because only the phase information is used. Experiment result （EER=0. 6%） on PolyU palmprint database illuminates the efficiency of the proposed approach.