Motion and Pose Estimation for Autonomy of Dynamic/Contact Tasks

Machine vision technology is a maturing field. This project seeks to move the technology from the theoretical and controlled-environment testing to a prototype stage for use in a space environment with a cluttered field and dynamic effects. In the end, the prototype system will demonstrate operational feasibility of using machine vision to provide supervised automation for target alignment/grasping strategies.

This project seeks to advance motion and pose estimation in support of dexterous robotics operations. Existing edge detection and model-based feature tracking methods will be used as the starting point for evolving a more robust technique for use in a dynamic environment. The challenges to overcome are related to dynamic effects of manipulator motion and orbital lighting.