ntnu-arl/underwater-datasets

多相机水下视觉惯性数据集是由挪威科技大学（NTNU）自主机器人实验室收集的。这个数据集包含了来自多种船载传感器的数据，包括5个相机、1个惯性测量单元（IMU）、一个气压计以及遥控潜水器（ROV）的执行器命令。数据是在受控的室内设施和自然水下环境中收集的。数据集被组织成多个子集，每个子集对应一个不同的环境，每个子集目录包含多个轨迹目录，每个轨迹目录包含原始传感器数据的`.bag`文件和通过ReAqROVIO获得的参考轨迹的`.tum`文件。

The Multi-Camera Underwater Visual-Inertial Dataset is collected by the Autonomous Robots Lab, NTNU. This dataset includes data from a variety of onboard sensors, such as 5 cameras, 1 IMU, a barometer, and actuator commands from the ROV. The data was collected in both controlled indoor facilities and natural outdoor underwater environments. The dataset is organized into multiple subsets, each corresponding to a different environment, with each subset directory containing multiple trajectory directories that include a `.bag` file with raw sensor data and a `.tum` file with the reference trajectory obtained by running ReAqROVIO.