SubT-MRS
收藏资源简介:
SubT-MRS数据集是由卡内基梅隆大学开发的,旨在推动SLAM技术在各种天气环境下的应用。该数据集包含30个多样化的场景,如无结构的走廊、不同光照条件和感知障碍物如烟雾和尘埃。数据集使用多种传感器,包括激光雷达、鱼眼相机、惯性测量单元和热像仪,并涉及多种移动方式,如空中、腿部和轮式机器人。创建过程中,数据集涵盖了从2019年到2023年的数据,包括室内外设置,如长走廊、越野场景、隧道、洞穴、沙漠、森林和灌木丛。SubT-MRS数据集的应用领域广泛,旨在解决SLAM技术在复杂环境中的鲁棒性和准确性问题,为未来的SLAM研究提供了一个关键的基准。
The SubT-MRS dataset was developed by Carnegie Mellon University to promote the application of SLAM technology across various weather and environmental conditions. This dataset consists of 30 diverse scenarios, including unstructured corridors, environments with varying lighting, and perceptual obstacles such as smoke and dust. It utilizes multiple sensors including LiDAR, fisheye cameras, inertial measurement units (IMUs), and thermal imagers, and supports multiple mobile platforms such as aerial, legged, and wheeled robots. The dataset covers data collected from 2019 to 2023, encompassing both indoor and outdoor settings like long corridors, off-road terrains, tunnels, caves, deserts, forests, and shrublands. The SubT-MRS dataset has a wide range of application scenarios, aiming to address the robustness and accuracy issues of SLAM technology in complex environments, and serves as a key benchmark for future SLAM research.

- 1SubT-MRS Dataset: Pushing SLAM Towards All-weather Environments卡内基梅隆大学 · 2024年



