With the rapid development of machine vision-related technologies, machine vision has drawn great attention from researchers in other engineering areas. In fields like mechanical engineering, a visual system could play a significant role in solving problems for robots’ motion control. If the pose of a machine tool manipulator can be obtained easily and accurately, the control strategies for mechanisms would become more efficient. And this would highly promote newly designed machines into the real application. The trend of advanced manufacturing is that mechanical systems are moving towards full automation. However, during the transition, certain human involved processes of manufacturing still cannot be replaced completely. How to create a safe and efficient human-in-the-loop working environment with well-developed machine vision technologies would be one of the keys to the success of advanced manufacturing.

This study focuses on control schemes with the visual system for different types of mechatronics devices. This dissertation has been organized with three steps of studies. In the first step, we study how to promote well-designed machines for application. A vision-based motion control approach is proposed and validated for a recently designed generalized parallel robot. As we understand that it is hard to ignore that human-involved manufacturing is happening before the full automation, the second stage of this study is on how to design a control scheme for a robot to understand information from the surroundings and to be able to accomplish certain tasks by itself. A mobile robot motion control scheme is proposed and implemented in this study with the advanced visual technology to achieve the function of a certain moving target following. In the third part, the study focuses on an advanced Human-Robot Interaction with providing a control scheme for a mobile manipulator by signals from common hand gestures in real-time, which would contribute to the technology access ability in areas that robots has replaced intensive labor work but human experience is highly required. Each of these three parts has been present with a project that includes a control approach design for a specific robot system and the system validation experiment.