Autonomous Robotic Following Using Vision Based Techniques

The Intelligent Systems And Autonomous Controls (ISAAC) robot is an experimental autonomous research platform being developed to advance current dismount following applications. Specifically, vision based following using pedestrian detection. The current standard in mule applications is the following of GPS waypoints. ISAAC is designed to follow a specific person using solely vision based techniques. The core of the vision based algorithms used in this application is based on years of research from a collaboration of government and university partners. Stereo vision techniques determine a person, identify them as the leader, and map a path for autonomous following. This paper focuses on the initialization of these algorithms and the control scheme used to implement them on a real-world platform.