Indoor Multi Human Target Tracking Based on PIR Sensor Network

Abstract

In order to solve the problem of human target tracking in smart-home Wireless Sensor Network (WSN) environment, and only based on limited measurement data of Binary PIR sensors, the sensor networks joint likelihood is derived, which proposes the indoor PIR sensor network Binary Auxiliary Particle Filter (Bin-APF) fusion estimate algorithm further more. Meanwhile, as for the problem of multiple human targets measurement classification and trajectory association, combined with PIR’s binary measurement, an improved K-Nearest Neighbor algorithm is adopted. And according to parameters of current experimental environment, a simulation is carried out, which contributes to the algorithm proposed. Experiment and Simulation results indicate that the MTT-KNN-Bin-APF algorithm accord well with the expectation of in-home multiple human target localization and tracking in consideration of actual result and error precision. Moreover, the algorithm is in low dependency of sensor network’s layout, which is suitable for various type of household arrangement. The method provides a solution to indoor human target tracking and is promising in the field of smart home.