Human Head Tracking Based on Particle Swarm Optimization and Genetic Algorithm

Abstract

This paper compares particle swarm optimization and a genetic algorithm for perception by a partner robot. The robot requires visual perception to interact with human beings. It should basically extract moving objects using visual perception in interaction with human beings. To reduce computational cost and time consumption, we used differential extraction. We propose human head tracking for a partner robot using particle swarm optimization and a genetic algorithm. Experiments involving two maximum iteration numbers show that particle swarm optimization is more effective in solving this problem than genetic algorithm.