Finger-Articular Back Texture Recognition Based on Log Gabor

摘要

This paper presents a new biometrics pattern, finger-articular back texture recognition using Log Gabor wavelet. Firstly, hand back image was captured by special imaging device. Then joint texture was segmented, and located by a sliding window. The texture features were extracted based on the Log Gabor wavelet transform and were classified by the corresponding feature encoding classifier. The experiment results shown that finger-articular back texture can be used for personal authentication, Log Gabor wavelet is better than Gabor wavelet for pattern extraction, and the equal right rate was 98.21% in Finger-articular back texture recognition.