无人机视觉识别数据集

<p>AU-AIR无人机视觉识别数据集</p><p>数据来源：公开网站</p><p>数据特点：无人机捕获的多模式航空数据集。具有视觉数据，对象注释和飞行数据 (时间，GPS，高度，IMU传感器数据，速度)。AU-AIR 数据集是首个用于目标检测的多模态无人机数据集。它满足了无人机视觉和机器人领域的需求，拥有来自不同机载传感器的多模态数据，并推动了针对自主空中监控的计算机视觉和机器人算法的发展。</p><p>AU-AIR 具有以下特点：</p><p><br></p><p><br></p><ul><li>空中图像中的目标检测</li><li>&gt;2 小时原始视频</li><li>32,823 标注帧</li><li>132,034 目标实例</li><li>8个与交通监控相关的对象类别</li><li>帧还标注了时间、GPS、IMU、高度和 UAV 的线速度</li></ul><p>应用场景：无人机测绘、建筑检查等领域</p>

<p>AU-AIR UAV Visual Recognition Dataset</p><p>Data Source: Publicly available websites</p><p>Data Characteristics: A multimodal aerial dataset captured by UAVs. It includes visual data, object annotations and flight data (time, GPS, altitude, IMU sensor data, speed). The AU-AIR dataset is the first multimodal UAV dataset for object detection. It meets the requirements of UAV vision and robotics fields, provides multimodal data from different on-board sensors, and promotes the development of computer vision and robotics algorithms for autonomous aerial surveillance.</p><p>AU-AIR has the following characteristics:</p><p><br></p><p><br></p><ul><li>Object detection in aerial images</li><li>Over 2 hours of raw video</li><li>32,823 annotated frames</li><li>132,034 object instances</li><li>8 object categories related to traffic surveillance</li><li>Frames are also annotated with time, GPS, IMU, altitude and the linear velocity of the UAV</li></ul><p>Application Scenarios: Fields including UAV mapping and building inspection</p>