基于Log Gabor滤波的指纹纹理匹配*

分析了Log Gabor滤波器的性能，详述了用于指纹识别的Log Gabor滤波器的构造方法，在此基础上提出了基于Log Gabor滤波器的指纹纹理匹配算法。首先采用了一种快速有效的参考点定位方法，在确定有效区域并归一化后，通过傅里叶变换把指纹图像转换到频域，在频域进行Log Gabor滤波，最后在滤波图像中提取特征,并与传统方法作了比较。实验结果表明，所提出算法的性能优于基于Gabor滤波的纹理匹配方法和基于细节点的方法，提高了指纹识别的准确率。

Fingerprint texture matching based on Log Gabor filtering

This paper first described the features and designed method of Log Gabor filter in detail. On the basis, proposed a novel method for fingerprint texture matching, which was based on Log Gabor filtering. This method first utilized a quick and effective method of reference point location, then normalized the region of interest. To implement the Log Gabor filtering in frequency domain, the Fourier transform was utilized to transfrom fingerprint image into frequency domain. The last step was extracting feature. The performance was compared with the traditional method. Experiment results show that the proposed algorithm performs better than Gabor-based and minutiae-based method with promoting high reliability of fingerprint identification.