Ear Recognition Based on Gabor-SIFT

Abstract

Scale invariant feature transform is a local point features extraction method. It can find those feature vectors in different scale space which are invariant for scale changes and rotations, and are flexible for illumination variations and affine transformations. The paper chooses SIFT to extract key points of ear images. Then the features of key points are extracted with the local multi-scale analysis feature of the Gabor wavelet. In this way, every key point is represented by a series of multi-scale and multi-orientation Gabor filter coefficients. Finally Ear recognition based on these feature is carried out with Euclidean distance as similarity measurement. Experimental results show that proposed method can effectively extract ear feature points, and obtain high recognition rate by using few feature points. It is robust to rigid changes, illumination and rotations changes of ear image, provides a new approach to the research for ear recognition.