In this paper a vision-guided technology is applied to our newly developed wheelchair robot for autonomous stair-climbing. In order to climb a flight of stairs automatically, a CCD system combining with a line laser mounted on this wheelchair robot is used to detect stairs. After performing the detection on the edges of steps and image processing, the captured image coordinates can be transformed into the real-world coordinates, and the required dimensions of steps are deduced. The corresponding rotational angles for the associated arms of the wheelchair robot are then calculated from the kinematics and one-step-ahead motion planning algorithm such that the autonomous vision-guided stair-climbing can be implemented reliably. Finally, the experiments validate that the wheelchair robot can climb stairs autonomously using the vision-guided technology.
展开▼
