DroneVehicle 大规模无人机航拍车辆检测数据集

这个数据集是天津大学的研究团队在进行无人机航拍图像的车辆检测和计数研究过程中收集和标注的。研究团队于 2020 年发布，相关论文成果为「Drone-based RGB-Infrared Cross-Modality Vehicle Detection via Uncertainty-Aware Learning」。

This dataset was collected and annotated by a research team from Tianjin University during their study on vehicle detection and counting using unmanned aerial vehicle (UAV) aerial images. The team released this dataset in 2020, alongside the corresponding academic paper titled "Drone-based RGB-Infrared Cross-Modality Vehicle Detection via Uncertainty-Aware Learning".

搜集汇总

背景与挑战

背景概述

DroneVehicle是一个大规模无人机航拍车辆检测数据集，由天津大学于2020年发布，包含56,878幅RGB和红外图像，提供多模态数据支持。数据集对五种车辆类别进行了方向性边界框的丰富标注，适用于目标检测和遥感应用，图像经过预处理可调整尺寸以优化训练效果。