Tracking with constraints in a web of sensors

Abstract

With the dramatic fall in price of electronics over the past several years, large-scale networks of sensors are steadily becoming more feasible. The goal of this research project was to deploy a sensor network for data collection of human trajectories and to develop and test a method for taking walls into account using a penalty function in an ongoing research project in trajectory tracking model. There were two deployed sensor networks, one with distance sensors and one with cameras, and the camera network was used to collect data for the two papers. The trajectory tracking model was modified to incorporate wall constraints to enable exploration of more realistic scenarios, with encouraging preliminary results.