M2UD
收藏资源简介:
M2UD数据集是由中国科学院沈阳自动化研究所等多个机构联合创建的多模态、多场景、不平地形机器人SLAM数据集。该数据集包含了58个序列,跨越12个类别,涵盖了城市、乡村、开阔地、长走廊、广场、地下停车场等多样化的挑战性环境,并在极端天气条件下进行了数据收集。数据集提供了通过RTK获得的平滑地面真实定位数据,以及用于定位和映射算法开发和评估的两个代表性场景的高精度毫米级激光扫描地图。
The M2UD dataset is a multimodal, multi-scenario, uneven terrain robotic SLAM dataset jointly created by the Shenyang Institute of Automation, Chinese Academy of Sciences and multiple other institutions. This dataset comprises 58 sequences spanning 12 categories, covering a wide range of challenging environments including urban areas, rural regions, open fields, long corridors, squares, underground parking garages and more. Data collection was conducted under extreme weather conditions. The dataset provides smoothed ground-truth localization data obtained via RTK, alongside high-precision millimeter-level laser scan maps of two representative scenarios for the development and evaluation of localization and mapping algorithms.




