Vision-based sensor fusion for human-computer interaction
S. Grange, E. Casanova, Terrence Fong, Charles Baur · 2003
This paper describes the development of efficient computer vision techniques for human-computer interaction. Our approach combines range and color information to achieve efficient, robust tracking using consumer-level computer hardware and cameras. In this paper, we present the design of the Human Oriented Tracking (HOT) library, present our initial results, and describe our current efforts to improve HOT's performance through model-based tracking.