Evaluating the Feasibility of a Particle Filtering Approach for Tracking Unmanned Surface Vehicles
Kishore Pamadi · 2009
A particle filter ing approach for tracking Unmanned Surface Vehicle s (USV s) with the use of surface radar is presented. This is a nonlinear estimation problem and involves non -Gaussian measurement noise because of glint in a radar problem. Glint is modeled as a Gaussian mixture comprising of two Gaussian distributions . Experimental data from tracking a Rigid Hull Inflatable Boat (RHIB) has been used to evaluate the Sampling Importance Resa mpling (SIR) filter and results have been compared with those obtained from applying a standard Extended Kalman Filter (EKF) .