Ear recognition based on Gabor scale information
Baoqing Zhang, Zhichun Mu, Zeng Hui, Hongbo Huang · 2013
As a promising biometrics, ear recognition is attracting increasing research interests among researchers in recent years. It has a wide range of civilian and law-enforcement applications. In this paper, a new feature extraction approach is investigated for ear recognition by using scale information of multi-scale Gabor filters. Compared with augmented Gabor features defined via concatenation of the Gabor filtering coefficients, the proposed Gabor scale feature will not only avoid too much redundancy but also tend to extract more precise structural information. So, the proposed feature is more robust to ear image variations. Rigorous experimental results on the ear image dataset of UND and USTB database III show the effectiveness of the proposed Gabor scale feature for ear recognition.