Human tracking system based on PIR sensor network and video

Abstract

To detect and locate the human target motion precisely, this paper intends to present a tracking algorithm based on pyroelectric sensor network and video analysis technologies. According to the advantages of pyroelectric sensor network system and video system, this paper uses weighted least squares to fuse the data which is collected by multiple heterogeneous sensor nodes to realize human target real-time tracking. Moreover, the data can also be collected by pyroelectric sensor network system and video cameras. Through simulation, the error of tracking results is analyzed. The results show that the method using homogeneous and heterogeneous sensors to fuse the measured vector obtain better human target real-time tracking effect. © 2014 Springer-Verlag Berlin Heidelberg.