Research of Manipulator Motion Planning Algorithm Based on Vision

摘要

The manipulator based on vision is used widely in many situations. However, there are many questions in the real-time control of manipulator based on vision, for example, the image processing is not exact, the image recognition is not fast and the process of the manipulator movement is not smoothly. All of these problems badly influence on the application of robots. In view of these drawbacks, this paper aims to present an analysis of various algorithms in the processing of manipulator based on vision in the control system, including: image recognition, image positioning motion planning and so on. Some improved solutions in the visual manipulator control system are presented to obtain better effective images and quickly response in the real-time system of eye and hand. The control system is fused eye-to-hand and eye-in-hand together by hybrid configuration.