The NavINST Dataset for Multi-Sensor Autonomous Navigation

This is a comprehensive multi-sensory dataset from various road trajectories in Kingston, ON, and Calgary, AB, Canada. This dataset features diverse lighting conditions and includes data from two indoor garages, with corresponding 3D maps. The dataset includes data from multiple commercial-grade IMUs and a tactical-grade IMU, along with a wide range of perception sensors. Notably, it is one of the first datasets to feature a solid-state LiDAR, alongside a mechanical LiDAR, four electronically scanning RADARs, a monocular camera, and two stereo cameras. Additionally, the dataset contains forward speed measurements from the vehicle's odometer and accurately post-processed high-end GNSS/IMU data, providing precise ground truth positioning and navigation information. The NavINST dataset is designed to facilitate advanced research in high-precision positioning, navigation, mapping, computer vision, and multi-sensor fusion, offering rich data for developing and validating robust autonomous vehicle algorithms. Finally, the dataset is fully integrated with ROS to ensure ease of use and accessibility for the research community. It also includes a toolkit for users working offline.

本数据集为一套全面的多感官数据集，源自加拿大安大略省金斯顿和阿尔伯塔省卡尔加里的多种道路轨迹。该数据集涵盖多样化的光照条件，并包含两个室内车库的数据，以及相应的三维地图。数据集汇集了多款商业级惯性测量单元（IMU）和一款战术级IMU的数据，以及广泛的感知传感器。尤为值得一提的是，本数据集是首批引入固态激光雷达（solid-state LiDAR）与机械式激光雷达并存的数据集之一，同时配备四部电子扫描雷达（RADAR）、一部单目相机和两部立体相机。此外，数据集还包含了车辆里程计的向前速度测量值，以及经过精确后处理的顶级全球导航卫星系统/惯性测量单元（GNSS/IMU）数据，提供了精确的地面实况定位和导航信息。NavINST数据集旨在促进高精度定位、导航、制图、计算机视觉和多传感器融合等领域的高级研究，为开发与验证稳健的自动驾驶算法提供了丰富的数据资源。最后，该数据集与ROS（机器人操作系统）完全集成，以确保其易用性和对研究社区的开放性。同时，还包括了一套适用于离线工作的用户工具包。