Intelligent gaze control for vision-guided humanoid walking: methodological aspects

Abstract This article presents methodological aspects of a scheme for visually guided humanoid robot navigation. The proposed approach is based on the maximization of the predicted visual information. For the information management a coupled hybrid Extended Kalman Filter is employed. Specific view direction evaluation strategies for conflicting objectives of different nature such as obstacle avoidance and self-localization, have to be weighted and pursued in parallel. A combination of both objectives shows the task dependence of the gazing strategy. A major goal of this work is to formalize and implement a decision making strategy in order to achieve an intelligent task-oriented active vision system for a biped walking robot. Simulation results based on experimental experience with real biped robots demonstrate the relevance of the approach.