Advanced face recognition system using Fourier Optics and Neural Networks

Abstract

Conventional pattern recognition uses wavelet method which can identify facial pattern but it is restricted to recognition of unmasked and fully exposed faces. Added this process is more time consuming. To overcome these limitations this system is designed in the light of Fourier Optics and Neural Networks. The system uses an advanced optical Fourier plane correlator for (real time) face recognition and feature extraction. Photographs are captured in real time and are applied as input to the system, which are compared with images from database in near light speed by the Fourier plane correlator. Applications include Airport security, Rail road security, Security at Malls and other places where people congregate together.