Swarm Optmization Algorithms for Face Recognition

Abstract

In this thesis, a face recognition system based on swarm intelligence is developed. Swarm intelligence can be defined as the collective intelligence that emerges from a group of simple entities; these agents enter into interactions, sense and change their environment locally. A typical system for face recognition consists of three stages: feature extraction, feature selection and classification. Two approaches are explored. First, Bacterial Foraging Optimization(BFO), in which the features extracted from Principal Component Analysis(PCA) and Linear Discriminant Analysis(LDA) are optimized. Second, Particle Swarm Optimization(PSO), which optimizes the transform coefficients obtained from the Discrete Cosine Transform(DCT) of the images. PCA, LDA and DCT are all appearance-based methods of feature extraction. PCA and LDA are based on global appearance whereas DCT is performed on a block by block basis exploring the local appearance-based features. Finally, for classification Euclidean distance metric is used. The algorithms that have been applied are tested on Yale Face Database.