A Model for Visual Observation under Uncertainty *

We address the problem of observing a moving agent. We advocate a modeling approach for the vi-sual system and its observer, where a discrete event dynamic system (DEDS) frame work is developed and “events ” are defined as ranges on parameter subsets. In particular, we propose a system for observing a manipulation process, where a robot liaiid manipu-lates an object. We recognize the liand/object in-teraction over time and a st.abilizing observer is coii-structed. Low-level modules are developed for recog-nizin the events tliat causes s1.at.e transitions wit Iiin the cfynamic manipulation system. The work esam-ines closely the possibilit irs for errors. mistakes and uncertainties in the manipulation system, observer construction process and rveiit ideiit ificatioii iiiecli-anlsms. The system utilizes different tracking tecli-niques in order to observe and recognize the t,ask i i i an acirue, adapiive and gonl-d i r rctcd nianner.