Moving Object Detection & Scenelock Tracking System

This book focuses on the identification, analysis and implementation of moving object detection and tracking algorithm using spatio-temporal frequency information of the image sequences on an embedded DSP platform. Real and virtual image datäs with static and dynamic environment are taken as test sequences and the movements with respect to these sequences has been analyzed, compared and optimized with respect to phase based and energy based algorithm using MATLAB. Based on quantitative analysis, energy based method was found to be better than phase based approach. Multiple objects detection and tracking system with energy based approach has been implemented using C and further optimized to suit for an Embedded DSP processor. It has been found that the computational time on Linux platform is better in comparison with MATLAB and embedded DSP processor. As a future work, it would be better to go for a high end embedded DSP processor to enhance the computation time.