Efficient face recognition method using RBF kernel and genetic algorithm

摘要

There area wider range of biometric authentication systems available; Face recognition system is one of them. It is an effective way for authentication and also considers many security aspects. Any face recognition system has to handle higher amount and dimension of image data. Whenever we consider this system at global level a large variety of problems frequently arrives. To overcome these problems this paper introduces a new model. This newly designed system uses radial basis function (RBF) kernel for management of small training sets of high dimension images with genetic algorithm based weight optimization technique. This system is effective for large datasets thus it uses genetic algorithm. Genetic algorithm provides fast learning and trains RBF neural network effectively, it reduces searching efforts and significantly reduces the time taken for recognition. This system is very effective at public places due to huge rush of public and provides fast and accurate recognition.