Data Repository from the Swarm of UAVs Innovate UK Project, Future Flights Strand 3, UAV Flights Dataset

This repository contains flight data captured by a UAV (Unmanned Aerial Vehicle) during landing approaches in both real-world and simulated environments. The data comprises video recordings in '*.mp4' format, viewable using media players like 'VLC media player' or 'Windows Media Player'. The footage was acquired using onboard cameras mounted on the UAV during real flights and recorded simulation flights using X-Plane 11.The recordings showcase UAV operations during the critical phases of landing, emphasizing vision-based navigation techniques for runway detection and trajectory evaluation. The dataset captures various scenarios encountered during real and simulated landing approaches, including diverse environmental conditions and flight parameters.The UAV flight data is instrumental for research and development activities focused on vision-based navigation systems for autonomous aircraft operations. Specifically, the dataset is intended to support tasks such as object detection, segmentation, and decision-making algorithms tailored for UAV landing approach and runway localization.Researchers and developers can leverage this dataset for reproducible experiments and algorithm validation, aiding in the advancement of technologies related to autonomous landing systems and UAV safety protocols. Importantly, the dataset is shared with explicit consent from all relevant stakeholders involved in the UAV operations depicted in the recordings.The Ground Truth folder contains the ground truth labels in csv foormat used for the side-lines evaluation of the vision-navigation system. Real and Simulation videos were used for the evaluation, describing the two side-lines of the runway. The videos used as ground truth are included in the real and simulation video repositories. The name of the video can be found in the first column of each .csv file.Please cite the related article when using the data repository.Tsapparellas, K., Jelev, N., Waters, J., Brunswicker, S., & Mihaylova, L.S., Vision-based Runway Detection and Landing for Unmanned Aerial Vehicle Enhanced Autonomy. 2023 In Proceedings of the IEEE International Conference on Mechatronics and Automation, ICMA, pp. 239-246, 2023.

本存储库收录了无人机（Unmanned Aerial Vehicle）在现实世界与模拟环境中着陆过程中的飞行数据。数据包括 '*.mp4' 格式的视频记录，可通过 'VLC 媒体播放器' 或 'Windows 媒体播放器' 等媒体播放器进行观看。这些视频资料由安装在无人机上的机载摄像头在真实飞行和利用 X-Plane 11 记录的模拟飞行中获取。视频资料展示了无人机在着陆关键阶段的操作，强调了基于视觉的导航技术在跑道检测和轨迹评估中的应用。数据集涵盖了现实和模拟着陆过程中遭遇的多种场景，包括多变的环境条件和飞行参数。无人机飞行数据对于专注于基于视觉的导航系统在自主飞行器操作中的研发活动至关重要。具体而言，本数据集旨在支持针对无人机着陆过程和跑道定位的物体检测、分割以及决策算法等任务。研究人员和开发者可以利用此数据集进行可重复实验和算法验证，助力于自主着陆系统和无人机安全协议相关技术的进步。值得注意的是，本数据集的共享得到了在记录中展示无人机操作的所有相关利益相关者的明确许可。《Ground Truth》文件夹包含用于视觉导航系统侧线评估的csv格式的真实标签。真实和模拟视频均被用于评估，描述了跑道的两侧线。作为真实标签使用的视频包含在真实和模拟视频存储库中。视频名称位于每个.csv文件的第一列。在使用数据存储库时，请引用相关文章。Tsapparellas, K.，Jelev, N.，Waters, J.，Brunswicker, S.，& Mihaylova, L.S.，基于视觉的跑道检测与无人机的着陆增强自主性。2023年，IEEE国际机电一体化与自动化会议（ICMA）论文集，第239-246页，2023年。