Target tracking using dynamic programming: algorithm and performance

Abstract

This paper presents a simple target tracking algorithm based on a dynamic programming approach. This is a 'Track-Before-Detect' technique, since no threshold is applied to the data at each measurement frame. Instead, data is processed over a number of frames before decisions on target existence are made. The track is returned simultaneously with the detection. We demonstrate the algorithm via simulations, and then analyse the detection and tracking performance of the algorithm, highlighting some problems with track before detect techniques in general.