Lightweight Signal Processing Algorithms for Human Activity Monitoring using Dual PIR-sensor Nodes

A dual Pyroelectric InfraRed (PIR) sensor node is used for human activity monitoring by using simple data processing techniques. We first point out the limitations of existing approaches, employing PIR sensors, for activity monitoring. We study the spectral characteristics of the sensor data for the cases of varying distance between the sensor and moving object as well as the speed of the object under observation. The sampled data from two PIR sensors, is first processed individually to determine the activity window size, which is then fed to a simple algorithm to determine direction of motion. We also claim that human count can be obtained for special scenarios. Preliminary results of our experimentation show the effectiveness of the simple algorithm proposed and give us an avenue for estimating more involved parameters used for speed and localization.