General purpose real object tracking using particle optimization algorithm

Abstract

Object Tracking is a field of computer vision which concerns with following and locating moving desired objects in video sequences . Previous techniques depend on detecting the entire image , pixel by pixel which is time consuming . Recently new biologically inspired approaches were developed that rely on Meta heuristic algorithms , such as Particle Swarm Optimization ( PSO ) algorithm , which proved to be robust and very fast . In this project we study the performance of the PSO algorithm to build a system for general purpose real time object tracking in dynamic environment and we apply it to indoor and outdoor environment . The PSO algorithm is used in each frame to search for the desired object and then to track the object in the video sequence . Color histogram is used to model the object to be tracked . The Hue color component demonstrates robustness under different illumination . The experimental results show that applying PSO algorithm to the object tracking field leads to efficient results . The PSO converges to desired object within seven iterations and it keeps tracking the object under different conditions