Robotic exploration under the controlled active vision framework

Robust techniques are presented for the derivation of depth from feature points on a target's surface and for the accurate and high-speed tracking of moving targets. These techniques are incorporated into a system that operates with little or no a priori knowledge of the object-related parameters present in the environment. The system is designed under the Controlled Active Vision framework and robustly determine parameters essential for performing higher level tasks such as inspection, exploration, tracking, grasping and collision-free motion planning. The system was implemented on the Minnesota Robotic Visual Tracker, a visual sensor mounted on the end-effector of a robotic manipulator combined with a real-time vision system.