Abstract
Ground segmentation using LiDAR technology plays a vital role in the successful execution of several tasks in agricultural robotics, such as sowing, spraying, fertilizing, harvesting, and weeding. However, traditional ground segmentation algorithms are often designed for urban environments and are not suitable for complex and challenging agricultural field environments. Additionally, some of these algorithms depend on specific LiDAR sensors, limiting the range of options available for use in agricultural robots. To address these limitations, we introduce FGSeg, a ground segmentation algorithm designed specifically for the agricultural field environment. Our proposed method utilizes only the spatial features of the point cloud data, making it compatible with a wide range of LiDAR sensors. Additionally, FGSeg can effectively distinguish between horizontal and slope terrains, which is crucial for many agricultural operations. The results of extensive experiments demonstrate that our proposed algorithm outperforms existing ground segmentation algorithms in both field and urban environments, and its real-time performance makes it well-suited for practical applications in the agriculture industry.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call