CMU-GPR
收藏资源简介:
CMU-GPR数据集是由卡内基梅隆大学机器人学院创建的,旨在研究地下辅助感知在机器人导航中的应用。该数据集包含15个轨迹序列,采集自三个无GPS的室内环境,包括地下测量数据、轮式编码器、RGB相机和惯性测量单元数据。数据集的创建过程涉及使用定制的SuperVision平台进行数据收集,该平台配备了多种传感器以确保数据的准确性和丰富性。CMU-GPR数据集主要用于机器人定位,特别是在无GPS环境下,通过地下信息进行定位,解决环境变化对视觉传感器的影响问题,适用于矿山、隧道等复杂环境下的机器人导航。
The CMU-GPR dataset was developed by the Robotics Institute of Carnegie Mellon University, with the goal of researching the application of underground auxiliary perception in robotic navigation. This dataset includes 15 trajectory sequences collected across three GPS-denied indoor environments, covering underground survey data, wheel encoder readings, RGB camera data, and inertial measurement unit (IMU) data. The data collection was carried out using a custom SuperVision platform, which integrates multiple sensors to guarantee the accuracy and richness of the collected data. The CMU-GPR dataset is mainly intended for robotic localization tasks, particularly in GPS-denied environments, where underground information is leveraged to perform localization and mitigate the impact of environmental variations on visual sensors. It is suitable for robotic navigation in complex scenarios such as mines and tunnels.




