drone-navigation-event-camera
收藏资源简介:
Edged-USLAM数据集是一个用于无人机导航与视觉SLAM研究的数据集,主要基于事件相机(DAVIS346)技术。该数据集包含同步记录的事件相机数据、IMU数据和运动捕捉地面真值数据,旨在评估Edged-USLAM算法的性能。数据集涵盖了多种运动轨迹(如直线、方形、激进转弯等)和不同光照条件(低光、HDR、动态光照等),以测试算法的几何和光度鲁棒性。数据集分为两个主要类别:运动类别(motion/)和光照类别(Illumination/),其中光照类别进一步分为使用IR过滤镜头(filtered/)和默认镜头(unfiltered/)的子集。每个序列都包含详细的时间戳同步和地面真值信息。数据集还提供了传感器的校准参数,包括相机内参、畸变参数和外参(相机到IMU的变换)。
The Edged-USLAM dataset is a benchmark dataset for unmanned aerial vehicle (UAV) navigation and visual SLAM research, primarily based on event camera (DAVIS346) technology. This dataset contains synchronously recorded event camera data, IMU data, and motion capture ground truth data, aiming to evaluate the performance of the Edged-USLAM algorithm. It covers various motion trajectories (e.g., straight lines, square paths, aggressive turns, etc.) and different lighting conditions (e.g., low-light, HDR, dynamic lighting, etc.) to test the geometric and photometric robustness of the algorithm. The dataset is divided into two main categories: the motion category (motion/) and the illumination category (Illumination/). The illumination category is further subdivided into subsets using IR-filtered lenses (filtered/) and default lenses (unfiltered/). Each sequence contains detailed timestamp synchronization and ground truth information. The dataset also provides sensor calibration parameters, including camera intrinsic parameters, distortion parameters, and extrinsic parameters (the transformation from the camera to the IMU).



