October 17, 2000, 1:00 PM
Zoran Duric

Department of Computer Science
George Mason University

Detection, Location, and Tracking of People in Video Images

ABSTRACT

A new method of finding people in video images is presented. Detection is based on a novel background modeling and subtraction approach which uses both color and edge information. We introduce confidence maps -- gray-scale images whose intensity is a function of our confidence that a pixel has changed -- to fuse intermediate results and to represent the results of background subtraction. The latter is used to delineate a person's body by guiding contour collection to segment the person from the background. Background subtraction is used as a preprocessing step for a tracking system. Tracking is performed at three levels of abstraction: regions, people and groups. People are tracked through mutual occlusions as they form groups and separate from one another. Strong use is made of color information to disambiguate occlusions and to provide qualitative estimates of depth ordering and position during occlusion. Simple interactions with objects can also be detected. Results for both indoor and outdoor sequences are presented.