A Visual Based Robot Trajectory Teaching Method for Traditional Chinese Medical Moxibustion Therapy

Abstract

At present, Chinese traditional physiotherapy is more and more trusted and valued by people, among which moxibustion is a very effective treatment. In order to improve the efficiency of human-computer interaction and help physiotherapists to use robots to replace human power, this paper proposes a visual based robot arm trajectory teaching method. The physiotherapy staff only needs to use an optical marker to demonstrate the therapeutic trajectory on the patient's back, and the robotic arm can pick up the expected trajectory and repeat the same work. We use LK optical flow algorithm to obtain marker's motion trajectory, and then smoothen the trajectory through SG filter, which is more executable for robot arm. Finally, we design an experiment and simulate it on MATLAB to obtain the accuracy of the trajectory and the results is impressive.