LuMon
收藏资源简介:
LuMon是由中东技术大学团队开发的月球单目深度估计综合基准数据集,包含真实嫦娥三号任务的高质量立体视觉深度数据及CHERI暗模拟数据集。该数据集涵盖6个子集,包含合成数据、地球模拟环境和真实月球影像,总数据量未明确但覆盖多种分辨率(最高2352×1728像素)和深度范围(最大50米)。数据通过立体重建技术生成,并经过严格的亚像素级校正和度量验证。其核心应用是解决月球车在极端光照、无大气散射条件下的自主导航问题,为地外感知领域提供首个跨域评估标准。
LuMon is a comprehensive benchmark dataset for lunar monocular depth estimation, developed by the team at Middle East Technical University. It includes high-quality stereo vision depth data from the real Chang'e 3 mission and the CHERI dark simulation dataset. This dataset consists of 6 subsets, covering synthetic data, Earth-simulated environments, and real lunar imagery. The total dataset size is unspecified, yet it spans multiple resolutions (up to 2352×1728 pixels) and depth ranges (up to 50 meters). The data is generated via stereo reconstruction techniques and has undergone strict sub-pixel calibration and metric validation. Its core application is to address the autonomous navigation challenges of lunar rovers under extreme lighting and atmosphere-free conditions, providing the first cross-domain evaluation standard for the extraterrestrial perception field.
MetuLuMon
数据集名称
MetuLuMon
数据集地址
https://metulumon.github.io/
数据集描述
MetuLuMon.github.io




