Data Fusion from Multiple Passive Sensors for Multiple Shooter Localization via Assignment.

摘要

This work examines the extension of a recently developed single target acoustic localization system to multiple targets. Multiple passive sensors distributed in a region of space will detect the acoustic signal of multiple gunfire events and measure the bearing to the event as well as the time the signal was detected. In order to alleviate the data association problem, an assignment-based method was developed in order to associate measurements at a fusion center, which will provide estimates of the multiple shooter locations. The assignment based approach is shown to localize multiple closely spaced shooters with a high degree of accuracy.