Race Classification Based on Facial Features

Human is unique and is an ancestral gift. Biometric recognition system can use race in identifying the human globally with a particular identity. This paper proposes an algorithm for classification of people into three major races namely Caucasoid, Negroid and Mongoloid, which can easily be integrated into any face recognition system. Firstly, the proposed classification method extracts the distinct primary facial feature, skin color model, and the other secondary features such as forehead and lip of the given face to effectively classify the races. Extensive simulations on images obtained from Yale, FERET and other sources exhibit superior performance of the proposed method.