MARKERLESS ARTICULATED HUMAN MOTION TRACKING USING HIERARCHICAL MULTI-SWARM COOPERATIVE PARTICLE SWARM OPTIMIZATION

Markerless articulated human motion tracking is an emerging tield with potential applications in areas such as automatic smart security surveillance. medical rehabilitation. computer based animations in games and movie industries. The primary objective of markerless articulated human motion tracking is to automatically infer human pose. expressed in tenms of joint angles from a video stream (sequences of images). However. extracting the articulated human body motion from multi-view synchronized video stream is a dit1icult task due to the underlying multimodal and high dimensional estimation problem. The Particle Filtering (PF) algorithm is the most extensively used tor generative model based articulated human motion tracking. However. it suffers from ·curse of dimensionality' and the challenge of ·particle degeneracy'. Furthermore. PF algorithm requires manual initialization and needs a sequence-specific motion model. Most recently. the swarm-intelligence based PSO algorithm have been gaining momentum in this tield