Główne pojęcia
This study presents a comprehensive multi-sensor dataset designed to facilitate the development of advanced perception and mapping techniques for 3D mapping in challenging indoor and outdoor environments.
Streszczenie
The DIDLM dataset integrates data from a diverse array of sensors, including infrared cameras, depth cameras, LiDAR, and 4D millimeter-wave radar, to address the limitations of traditional single-sensor SLAM methods in diverse environmental conditions. The dataset covers a wide range of challenging scenarios, such as rainy, snowy, and bumpy road conditions, as well as indoor and outdoor settings, providing a realistic background environment with interactive robot data at different speeds.
The key highlights of the dataset include:
- Integration of infrared cameras, depth cameras, LiDAR, and 4D millimeter-wave radar to enhance perceptual capabilities in extreme conditions.
- Comprehensive data collection under various weather conditions (sunny, rainy, snowy), lighting conditions (day, night), and road surfaces (flat, bumpy) for both indoor and outdoor environments.
- Comparison of SLAM performance between standard and extreme conditions at the same locations, enabling the assessment of the impact of different environmental factors on diverse sensors.
- Rigorous testing of state-of-the-art SLAM algorithms, including laser-based and visual SLAM, to identify their strengths and weaknesses in different scenarios.
The DIDLM dataset aims to address the scarcity of data in special environments and foster the development of perception and mapping algorithms for extreme conditions, advancing intelligent mapping and perception capabilities.
Statystyki
The dataset contains approximately 2000 GB of image and point cloud data, which was subsequently converted to approximately 250 GB in .mcap format.
The dataset covers a total distance of 24.161 km across various scenarios, including roundabouts, squares, parking lots, pathways, slopes, and other environments.
The dataset includes data collected under sunny, rainy, and snowy conditions, both during the day and at night, and on flat and bumpy road surfaces.
Cytaty
"This pioneering dataset not only enriches the domain of SLAM but also extends its applicability to researchers dedicated to unraveling the intricacies of outdoor mapping within such exacting environmental contexts."
"Leveraging multi-sensor data including infrared, depth cameras, LiDAR, 4D millimeter-wave radar, and robot interactions, the dataset advances intelligent mapping and perception capabilities."