Multiple human location in a distributed binary pyroelectric infrared sensor network

This paper proposes a multi-human locating method for distributed wireless sensor network with binary pyroelectric infrared sensors. The uniformly deployed infrared sensor network consists of one sink node and nine sensor nodes, which can detect infrared information of moving human targets. An anti-logic bearing-crossing location and clustering algorithm is proposed to locate different targets. Firstly, dynamic virtual detection lines are generated based on the angular bisector of sensor's FOV(field of view) and all intersection points of these detection lines are primary measurement points. The location of multi-human targets can be achieved by first clustering the primary measurement points and then assigning these clusters to each target, which can simplify the assignment problem from multiple points to several clusters. Finally an anti-logic primary measurement points filtering method is used to get the location result of each target. Simulation and experimental results have shown that the measurement points can be obtained and assigned to different targets effectively, and our proposed location method can locate and track two human targets well.