Research on Global Path Planning in Air-Ground Collaborative Robotic Systems Based on the ORB-SLAM3 Algorithm

The air-ground collaborative robotic system has garnered significant attention due to its potential applications in complex scenarios such as search and rescue and extraterrestrial exploration. However, existing research encounters substantial challenges related to efficient and reliable path planning in large scale environments. In this study, we explore the creation of maps in unknown environments without GPS by unmanned aerial vehicles (UAVs) to assist unmanned ground vehicles (UGVs) in rapid and advanced global path planning. Our approach begins with a UAV equipped with a camera, using the ORB-SLAM3 algorithm to generate a sparse point cloud map. Next, we apply the Statistical Outlier Removal (SOR) algorithm to filter out point cloud noise. To obtain a map of navigable paths, we employ the Random Sample Consensus (RANSAC) algorithm to extract the point cloud information representing the traversable ground. Finally, the processed 3D point cloud map is converted into a 2D grid map, resulting in an obstacle map suitable for UGV global path planning. Additionally, we employ the A* algorithm for path planning, considering critical factors such as path width and gradient. The proposed system demonstrates significant application potential in unknown environment exploration. UAVs can pre-scan unfamiliar terrains to facilitate efficient UGV navigation. The results indicate that our integrated method significantly enhances the efficiency and reliability of path planning in collaborative robotic systems.