变结构陆空机器人自主跨域越障技术研究

Land and air robots have attracted much attention due to their flexible maneuverability. In order to better utilize the characteristic of cross-domain obstacle negotiation, based on the variable structure land-air robot, the depth image was binarized to extract the edge contour of the obstacle ahead, and the minimum rectangle matching method was used to model the obstacles to obtain the distance and height of the obstacles. Furthermore, the autonomous cross-domain obstacle negotiation strategy was proposed according to the obstacle detection results and obstacle negotiation performance of terrestrial locomotion. The experimental results demonstrate that the robot can detect the distance and height of the obstacles ahead accurately, and cross the obstacle to reach the target point autonomously. The obstacle detection method and autonomous cross-domain obstacle negotiation strategy proposed in this paper are effective.