2008 Poster Sessions : Tracking Multiple People in Cluttered Scenes with a Wireless Camera Network

Student Name : Kyle Heath
Advisor : Leo Guibas
Research Areas: Graphics/HCI
Abstract
We describe a distributed vision-based technique for tracking multiple people with multiple cameras in dynamic and cluttered scenes. The technique is designed for environments where background modeling is difficult and occlusions occur in all views. Multiple stereo sensors individually estimate the 3D trajectories of salient feature points. Sensors communicate their sparse 3D measurements to other sensors with overlapping views. Each sensor fuses 3D measurements with a particle filter to track people in a common world coordinate frame. We describe a real-time implementation using PCs and inexpensive camera and evaluate its performance using the MOTA-MOTP multi-target tracking
performance metrics.

Bio
Kyle Heath is a currently a Ph.D. student in the department of Electrical Engineering at Stanford University. He received a B.S. degree in Computer Engineering from Rose-Hulman Institute of Technology in 2004 and a M.S. degree in Electrical Engineering from Stanford University in 2007.