Visual perception for human-robot manipulator coordination

• Main Author: Tan, Runjia
• Other Authors: Hu Guoqiang
• Format: Thesis-Master by Coursework
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Electrical and electronic engineering::Control and instrumentation::Robotics; Engineering::Computer science and engineering::Computing methodologies::Image processing and computer vision
• Online Access: https://hdl.handle.net/10356/155616
• Institution: Nanyang Technological University

Human-robot coordination or HRI have a wide range of applications in autonomous vehicles, logistics planning, and cooperative inspection, etc. For the perception module widely used in HRI, such as vision, emotion, chatting and action, the visual sensors can perform well in many fields due to their ease of use and mature processing algorithms. This dissertation mainly focuses on the introduction and our implementations of Human-Robot Interaction (HRI) system based on visual perception. Firstly, the introduction part gives the definition of HRI and a glance of our experiment, that is, human-to-robot (H2R) handover. Then, we list and introduce some related works that inspires us from three aspects: the overall system, perception module, especially for grasping, and control system due to the demand of visual servo. Then, according to analysis and follow for previous work, we present a system for H2R handover with arbitrary objects step-by-step. This part includes three sub-parts as perception, trajectory planning and control system. Finally, we will combine all these parts to compose a complete system. And the robot can grasp some common objects but never seen before from human hand while avoiding collision between robot arm and human body successfully.