Human feet tracking in arranging the navigation of a robotic rollator

This paper presents an indirect user interface solution for an active walking aid using visual image sensing. The aim of the design is to track the movement of the human feet in order to obtain command signals for driving an active walker to be seen as a robotic rollator. The human gait becomes the input device by which the user controls the actuators of the robotic rollator. Computer vision techniques are used to track the user's feet as end effectors during human gait. A camera is mounted on the frame of the rollator and captures video images of the user's feet. The kinematics of the feet are measured in two dimensions to obtain the position and orientation of the user's body. The measured motion parameters of the feet are relayed to the desired position and orientation of the rollator. The robotic rollator is to be used as a rehabilitation tool during everyday walking and gait training.