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

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