A New Hybrid Method For Face Recognition

Abstract

Now a days, security is a major concern for any organization. It is very difficult to have enough faith in any person as far as security of the organization is concerned. Due to these reasons, face recognition gets popularity in the security domain. Many conventional methods are available to do the face recognition. In this paper, we have discussed few of them covering advantages, disadvantages and applications. It is not possible to have a single face recognition method to cover all underlying applications of face recognition system. We have also device a new hybrid method by combining existing approaches of Local Binary Pattern (LBP) and Histogram. This hybrid approach has been tested on standard dataset and compared with LBP. Simulation results shows that proposed hybrid approach outperforms compared to LBP as far as security and speed is concerned.