A Visual Search with Pick-up Balls Robot through Path Planning in Known Environments

Autor: Chien-hsin Wu, 吳建欣
Rok vydání: 2009
Druh dokumentu: 學位論文 ; thesis
Popis: 98
The purpose of this paper is to achieve a visual search robot system which has webcam on it. The robot searched and picked up the targets by visual servo control. It also owns the capability of the path planning, and completed the pick-up job in known environment effectively. In terms of detecting the targets, the most important problem is how to find out the targets in the image of the grabbed stream video. In this paper, we take the table-tennis ball for the target. The robot detected the table-tennis balls by template matching, and got the distance and the location of the balls. Finally, we control the robot to pick up the balls. In terms of path planning, we improve the completed coverage path planning to suit our robot system in this paper. By this, the robot can work more effectively in table-tennis training room. On the other hand, we design a special mechanism, and it can pick up the table-tennis balls easily without large energy. After experiments, our system can quickly pick up the table-tennis balls on the ground, and also can dodge the table-tennis tables and the walls. Through the improved path planning and the special mechanism, the athletes will not pick up the balls hardly after training. Therefore, they can train easily and absorbedly.
Databáze: Networked Digital Library of Theses & Dissertations