Particle Swarm Optimization Aided Kalman Filter for Object Tracking

Object tracking aims to detect the path of objects moving randomly by obtaining input from a series of images. Automatic detection and tracking of object is an interesting area of research for defence related applications like missile tracking, security systems and commercial fields like virtual reality interfaces, robot vision etc., Kalman filter tracks the object by assuming the initial state and noise covariance. For efficient tracking by any filter like Kalman filter noise covariances must be optimized. Here in this paper we propose tuning of noise covariances of Kalman filter for object tracking using particle swarm optimization (PSO). Here we consider not only object features but also object motion estimation to speed up the searching procedure. Experimental results of tracking a ball demonstrate that the proposed method is efficient under dynamic environment.