HOPS (hierarchical orchard panoptic segmentation)
收藏资源简介:
HOPS数据集是由德国波恩大学机器人中心创建的,包含真实苹果园中不同传感器(如地面激光扫描仪、无人机上的RGB-D相机等)收集的点云数据。该数据集特别为分层全景分割任务设计,记录了两年内不同生长阶段的苹果园数据,并提供了高质量的注释,包括语义分割、树木实例分割、果实和树干实例分割。HOPS数据集旨在支持精准农业中的作物产量估算,通过使机器人能够理解周围环境来识别目标对象。
The HOPS dataset was created by the Robotics Center of the University of Bonn, Germany. It contains point cloud data collected by various sensors in real apple orchards, such as terrestrial laser scanners and RGB-D cameras mounted on unmanned aerial vehicles (UAVs). Specifically designed for the hierarchical panoptic segmentation task, this dataset records orchard data across different growth stages over a two-year period, and provides high-quality annotations including semantic segmentation, tree instance segmentation, as well as instance segmentation of fruits and tree trunks. The HOPS dataset aims to support crop yield estimation in precision agriculture, enabling robots to understand their surrounding environment and identify target objects.

- 13D Hierarchical Panoptic Segmentation in Real Orchard Environments Across Different Sensors德国波恩大学机器人中心 · 2025年



