Neural Networks and 3D Edge Genetic Template Matching for Real-Time Face Detection and Recognition

Abstract

This chapter describes a novel system that can track and recognize faces in real time using neural networks and genetic algorithms. The main feature of this system is a 3D facemask that combined with a neural network based face detector and adaptive template matching using genetic algorithms, is capable of detecting and recognizing faces in real time. Neural network learning and template matching enable size and pose invariant face detection and recognition while the genetic algorithm optimizes the searching algorithms enabling real time usage of the system. It is hoped that this chapter will show how and why neural networks and genetic algorithms are well suited to solve complex pattern recognition problems like the one presented in this chapter.