This paper describes an end-to-end pipeline for tree diameter estimation based on semantic segmentation and lidar odometry and mapping. Accurate mapping of this type of environment is challenging since the ground and the trees are surrounded by leaves, thorns and vines, and the sensor typically experiences extreme motion. We propose a semantic feature based pose optimization that simultaneously refines the tree models while estimating the robot pose. The pipeline utilizes a custom virtual reality tool for labeling 3D scans that is used to train a semantic segmentation network. The masked point cloud is used to compute a trellis graph that identifies individual instances and extracts relevant features that are used by the SLAM module. We show that traditional lidar and image based methods fail in the forest environment on both Unmanned Aerial Vehicle (UAV) and hand-carry systems, while our method is more robust, scalable, and automatically generates tree diameter estimations.

No abstract

Mobile robots such as unmanned ground vehicles (UGVs) and unmanned aerial vehicles (UAVs) are increasingly used for precision agriculture. While UGVs have larger payload capabilities and longer operation time, they are limited to 2-D space. This makes UAVs better suited for tasks that require fast coverage, harsh terrain traversal, and high altitude or multilevel operation. However, it remains a challenging task to develop a reliable yet fully autonomous UAV system that can actively extract actionable information in large-scale cluttered agricultural environments. Such a system will have to estimate its own poses, build a map of the environment, navigate through obstacles, and act to gather information with limited onboard computation and battery life. In this survey, we first review recent advances in UAV hardware and software, ranging from novel platforms and sensors to state-of-the-art autonomous navigation, object detection and segmentation, robot localization, and mapping algorithms related to agriculture. We then provide a list of challenges in each field and potential opportunities for the broader adoption of UAVs in precision agriculture.

In this letter we present a novel descriptor based on polygons derived from Urquhart tessellations on the position of trees in a forest detected from lidar scans. We present a framework that leverages these polygons to generate a signature that is used detect previously seen observations even with partial overlap and different levels of noise while also inferring landmark correspondences to compute an affine transformation between observations. We run loop-closure experiments in simulation and real-world data map-merging from different flights of an Unmanned Aerial Vehicle (UAV) in a pine tree forest and show that our method outperforms state-of-the-art approaches in accuracy and robustness.