Indoor human tracking and state estimation by fusing environmental sensors and wearable sensors

In this paper, we aim to track a human's indoor location while estimating his/her behavioral state. We use environmental PIR sensors to localize the individual and assume that the position of the individual obeys the bivariate Gaussian distribution, which can be used to improve the tracking accuracy. At the same time we introduce a wearable acceleration sensor to estimate the human's state. The hardware setup consists of two types of sensor: passive infrared sensors and a three-axis acceleration sensor with built-in bluetooth communication. By comparing the tracking performance between the PIR sensor-based algorithm and the fusion-based algorithm, we find that the latter has less errors. This indoor localization system can be used in future smart homes.