Particle Filters for Tracking Target with a Camera

A solution for tracking the moving target using particle filters (Sequential Monte Carlo methods) is developed. It consists of a motion model of moving target according to the Newton´s law and a non-linear measurement equation in position. The measurements are taken by detecting the target in video frames which are acquired from a camera. A tracking algorithm is presented which uses a swarm of particles to estimate the possibility distribution of the motion states. This is of great importance for high-performance real-time applications such as public monitoring system and no-touching man-machine interaction systems etc.