Human detection and tracking system in controlled environment for collaborative surgical robot

Abstract

Robotic collaboration applied to surgery has great potential. It is estimated that an important challenge is the dynamic interaction of the robot in the surgical field, indicating the preference of systems that contribute to the improvement of cooperation between human and machine. With that in mind, we plan to implement an architectural vision system internally to the project "Collaborative robotics and neuronavigation applied to neurosurgery" in development at the Applied Technologies Laboratory (LTA) of the University of São Paulo, through Scientific Initiation. The study and development of software for detection and tracking in the space of people and their members, through conventional cameras and also with cameras with depth perception. The detection aims to search for the body parts in the image, while tracking seeks to estimate the coordinates of these targets in space. The motivation of this development is the improvement of human-robot interaction, as the evaluation the positions and movements of users in the vicinity of the robot.