基于Log Gabor滤波和不变矩的指纹识别

The identification of fingerprints based on Log Gabor filter and invariant moment

文中提出了基于八个方向的Log Gabor滤波和不变矩的指纹匹配算法,遵循找中心点、分扇区、归一化、八个方向的Log Gabor滤波、指纹特征提取、加权混合匹配算法的步骤,能较好的定位指纹,且能解决指纹平移变换的问题,最大可能的容忍指纹的部分缺失和关键点的遗漏,具有较好的容错率。此外,该指纹匹配算法具有准确性高,可移植性强的特点,使其可以在很多领域得到广泛应用。

This paper presents a modified algorithm based on the Log Gabor filtering within 8 directions and invariant moments. Following the main procedures of this algorithm such as centering, sectoring, feature extraction and mixed matching,it is able to locate the fingerprints accurately. It can also solve the problems resulted from the turning and movements of fingerprints. This algorithm accommodates the situation where there is loss of key points and part of the pattern. A respectively high error-tolerate rate also makes this kind of algorithm run well. What ’ s mor e,the high accuracy and transportability support its widely use in separated fields.