Bio Smart Secure

日期 : Sep 2017 - 现在

Designing effective face recognition system is a fundamental problem in computer vision. Over the last couple of years, various new techniques of face recognition have been developed. These developments are being stimulated by advancements in computer vision techniques which assures less error rate in face recognition systems. In Face recognition system, there is a major drawback that they are not efficient and accurate in real environment. We propose face recognition system which identify individuals out of a large database containing several images per person. Face is represented by featured matrix, based on Local tetra pattern(LTrP) which is then used to train ELM classifier. Local tetra patterns provide more information by encoding each image into four distinct values. ELM is a single layer (or in some cases multiple layers) of hidden neurons feed forward neural network for classification. The experimental results show that our proposed system improves the accuracy and performance of a face recognition system significantly.