A Novel Face Parts Detection Mechanism for Biometric Recognition

An efficient technique for detecting and recognizing face parts in surveillance systems

Abstract

In surveillance, the role of Closed-Circuit Television (CCTV) is found more significant in recent days for security reasons in different places such as town centres, airports, public transport, evidence in court etc. Human face detection is the most important part in the body that helps to identify the person during identification. This paper has proposed a new technique that accurately detects different parts of the face such as nose, eyes and mouth. During detection, this technique also considers the illumination and background changes of the facial parts. This work is composed of three phases. The first is the computation of an integral image representation which allows the features to be computed quickly. The second is a learning algorithm based on modified AdaBoost which selects a small number of critical visual features from a larger set and provides an efficient classifier. The third phase is a method for combining increasingly more complex classifiers. The proposed technique provides good results maintaining a low computational cost. Results obtained from experiments shows that the proposed technique gives better results than other state-of-the-art techniques.