A Multiple Hypothesis Tracker for a Distributed Network of Sensors

Abstract

This paper describes the multiple hypothesis network tracker (MHNT) being developed by BAE Systems. The goal here was to design, develop and implement a tracker capable of tracking multiple vehicles moving across a network of geographically distributed static sensors, where each sensor's coverage area is small compared to the distance between sensors. Each sensor reports the time when a target crosses the sensor's coverage area sensors may also report features associated with the target. A centralized tracker receives reports from all of the sensors and, using prior information regarding target motion across the network of sensors and actual travel times, assign reports to tracks and computes the most likely set of assignments