MOR-UAV
收藏资源简介:
MOR-UAV是一个大规模的无人机视频数据集,专注于移动对象识别(MOR)任务。该数据集由马拉维亚国家技术研究所视觉智能实验室创建,包含从30个无人机视频中收集的10,948帧,标注了89,783个移动对象实例。数据集涵盖多种复杂场景,如天气变化、遮挡、飞行高度变化和多视角摄像机。主要关注两类车辆:汽车和重型车辆。MOR-UAV旨在支持无人机视觉应用,如空中监视、搜索救援和城市乡村场景理解,解决实时视频流中移动对象的定位和分类问题。
MOR-UAV is a large-scale unmanned aerial vehicle (UAV) video dataset dedicated to the Moving Object Recognition (MOR) task. Developed by the Visual Intelligence Laboratory of Malaviya National Institute of Technology, this dataset contains 10,948 frames collected from 30 UAV videos, with 89,783 moving object instances annotated. It covers a variety of complex scenarios including weather variations, occlusions, changes in flight altitude, and multi-view camera setups. The dataset primarily focuses on two categories of vehicles: passenger cars and heavy-duty vehicles. MOR-UAV is designed to support UAV-based visual applications such as aerial surveillance, search and rescue, and urban and rural scene understanding, addressing the issues of localization and classification of moving objects in real-time video streams.
- 1MOR-UAV: A Benchmark Dataset and Baselines for Moving Object Recognition in UAV Videos马拉维亚国家技术研究所视觉智能实验室 · 2020年



