The TAIL-Plus dataset is an extension of the previous TAIL (Terrain-Aware MultI-ModaL) dataset, focusing on robot localization and mapping in planetary surface analog environments. The dataset was collected using both wheeled and quadruped robot platforms equipped with a diverse sensor suite, including 3D LiDAR, RGB-D cameras, global-shutter color cameras, RTK-GPS, and IMU.
The dataset features a wide range of conditions, including different types of locomotion, multi-loop trajectories, day-night illumination changes, and varying surface terrain characteristics (coarse and fine sand). These challenging scenarios are designed to test the robustness and accuracy of multi-sensor SLAM algorithms for field robots in unstructured, deformable granular environments, which are representative of planetary exploration tasks.
The dataset provides time-synchronized, spatially-calibrated sensor data, as well as 6-DOF ground truth poses from the IMU-integrated RTK-GPS system. This comprehensive dataset aims to support the development and evaluation of various multi-sensor SLAM approaches, such as LiDAR-inertial, visual-inertial, and LiDAR-visual-inertial fusion.
The authors plan to further expand the dataset by incorporating additional sensor modalities, such as event cameras, infrared thermal cameras, and solid-state LiDARs, as well as exploring more diverse and challenging environments for planetary exploration analog scenarios.
To Another Language
from source content
arxiv.org
Key Insights Distilled From
by Zirui Wang,C... at arxiv.org 04-23-2024
https://arxiv.org/pdf/2404.13600.pdfDeeper Inquiries