Ground multiple target tracking with a network of acoustic sensor arrays using PHD and CPHD filters

In this paper, we explore the potential of networked microphone arrays for multiple target tracking. Tracking is accomplished by using the direction-of-arrival (DOA) estimates of multiple microphone arrays. Each microphone array obtains the DOA estimates by using the wideband extensions of the multiple signal classification (MUSIC) technique. Based on these DOA estimates, multi target tracking is done by using Probability hypothesis density (PHD) and cardinalized probability hypothesis density (CPHD) algorithms. The results show that the CPHD performs better than the PHD on real data as it estimates the number of targets more accurately.