Robot Hand/Eye Control Algorithms Using 3D Machine Vision
| Autor: | I-Ping Chang, 張宜平 |
|---|---|
| Rok vydání: | 2006 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 94 This thesis proposes a binocular machine vision using image processing algorithms for controlling a robot grasping a moving target. Binocular cameras are combined with a robot in a hand-eye cooperative configuration using the cameras as vision feedback sensors for controlling the manipulators. The binocular cameras are used to grab sequential images in which the moving target is detected using subtraction, binary, projection, and cross-correlation image processing procedures. The geometric relations for standard stereo geo-metry (SSG) are used to calculate the trajectory of the m-oving target in world coordinates. The target trajectory is used for determining the inverse kinematic solutions for the robot joint command, thereby accomplishing visual servo control. An algorithm for searching the correspondence points in the left and right images is proposed. This method is expected to have fast 3D target trajectory calculation. The shortest path planning for the robot to grasp the moving object is also a research topic in this thesis. The proposed experimental system is implemented and experimental results are used to demonstrate that the proposed vision control algorithms are feasible. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|