AgriLiRa4D
收藏资源简介:
AgriLiRa4D是由拓普枪机器人、杭州电子科技大学及南京大学联合创建的多模态无人机数据集,专为挑战性户外农业环境设计。该数据集规模宏大,涵盖平坦、丘陵和梯田三种典型农田类型以及边界与覆盖两种作业模式,共包含六组飞行序列,同步采集了三维激光雷达、四维雷达和惯性测量单元的高精度数据,并辅以厘米级精度的光纤惯性导航系统实时动态测量真值轨迹。数据集通过精心校准的传感器套件在真实农田场景中系统采集,旨在支持鲁棒的同时定位与地图构建研究,重点解决农业无人机在低纹理作物、重复模式、动态植被等复杂环境下的自主导航与定位难题,推动精准农业自动化技术的发展。
AgriLiRa4D is a multimodal UAV dataset jointly created by Top Gun Robotics, Hangzhou Dianzi University and Nanjing University, designed specifically for challenging outdoor agricultural environments. The dataset has a substantial scale, covering three typical farmland types: flat, hilly and terraced fields, as well as two operation modes: boundary flight and coverage flight, and includes a total of six flight sequences. High-precision data from 3D LiDAR, 4D radar and inertial measurement unit (IMU) are synchronously collected, with real-time dynamic ground-truth trajectory measured by a centimeter-level accuracy fiber-optic inertial navigation system as supplementary reference. The dataset is systematically collected in real farmland scenarios using a carefully calibrated sensor suite, aiming to support robust simultaneous localization and mapping (SLAM) research, focusing on addressing the challenges of autonomous navigation and positioning for agricultural UAVs in complex environments such as low-texture crops, repetitive patterns and dynamic vegetation, so as to promote the development of precision agriculture automation technologies.
AgriLiRa4D数据集概述
数据集名称
AgriLiRa4D
核心描述
AgriLiRa4D是一个多传感器无人机数据集,旨在为具有挑战性的农业场景提供鲁棒的SLAM(同时定位与地图构建)解决方案。
引用信息
- 标题:AgriLiRa4D: A Multi-Sensor UAV Dataset for Robust SLAM in Challenging Agricultural Fields
- 作者:Zhihao Zhan, Yuhang Ming, Shaobin Li, Jie Yuan
- 年份:2025
- arXiv标识:eprint={2512.01753}, archivePrefix={arXiv}, primaryClass={cs.RO}
- 论文链接:https://arxiv.org/abs/2512.01753
许可信息
本网站内容遵循知识共享署名-相同方式共享4.0国际许可协议。

- 1AgriLiRa4D: A Multi-Sensor UAV Dataset for Robust SLAM in Challenging Agricultural Fields拓普枪机器人(南京)、杭州电子科技大学、南京大学 · 2025年



