Donald Bren School of Information and Computer Sciences

The focus of this study is to develop an understanding of the state of the art in visual motion perception by intelligent systems. We examine diverse theoretical and empirical approaches to visual motion analysis, perception, and understanding. We survey the literature applicable to visual motion perception as discussed in scene analysis techniques, hardware-based vision systems, computer animation, artificial intelligence techniques, and human motion perception. We focus primarily on issues of observed object and image sequence description, representation, and perceptual control strategies. We next introduce two concepts in visual motion perception: motion vantage perspective and object motion coherence. We then suggest the emerging trends, problem areas, and alternative directions for research in visual motion perception by intelligent systems.