Technical Validation Experiments for "MMFW-UAV Dataset: Multi-Sensor and Multi-View Fixed-Wing UAV Dataset For Air-to-Air Vision Tasks"

This dataset is a subset of the MMFW-UAV Dataset used for technical validation experiments in the paper "MMFW-UAV Dataset: Multi Sensor and Multi View Fixed Wing UAV Dataset For Air to Air Vision Tasks". Specifically, 3,000 images with different views were randomly selected from the image datasets collected by different types of sensors, and three experimental sub-datasets were created, i.e., Zoom-Sensor-Subset, Wide-Angle-Sensor-Subset, and Thermal-Imaging-Sensor-Subset. These subsets were further divided into training and validation sets according to the ratio of 8:1:1.本数据集为论文《MMFW-UAV数据集：面向空对空视觉任务的多传感器多视角固定翼无人机数据集》（MMFW-UAV Dataset: Multi Sensor and Multi View Fixed Wing UAV Dataset For Air to Air Vision Tasks）中用于技术验证实验的MMFW-UAV数据集（MMFW-UAV Dataset）的子集。具体而言，研究人员从不同类型传感器所采集的图像数据集内随机遴选3000张不同视角的图像，并构建了三个实验子数据集，即变焦传感器子数据集（Zoom-Sensor-Subset）、广角传感器子数据集（Wide-Angle-Sensor-Subset）与热成像传感器子数据集（Thermal-Imaging-Sensor-Subset）。随后按照8:1:1的比例将上述子数据集进一步划分为训练集与验证集。

搜集汇总

背景与挑战

背景概述

该数据集是MMFW-UAV数据集的子集，用于技术验证实验，包含从多传感器（变焦、广角和热成像）采集的3,000张多视角图像，划分为三个子集并按8:1:1比例分为训练和验证集，旨在支持空对空视觉任务的研究。