An Ear Identification System Using Local-Gabor Features and KNN Classifier

Biometric recognition systems are inquired to solve security problems in networked society recently. Uniqueness of ear shape and its robustness in spite of increasing age has attracted researcher´s attention. In this paper, an appropriate biometric system is presented based on local-Gabor features which effectively extract appropriate features bring along dimensionality reduction. Subsequently the KNN classifier is applied. Experimental results show the effectiveness of the proposed method which brings both higher recognition performance and higher speed benefits.