Indoor-Multi-Level Fog Dataset (IMFD)

IMFD是由泰德大学计算机工程系创建的多级雾化图像和视频数据集，旨在为去雾方法提供基准测试。该数据集包含两个真实场景的彩色视频数据，具有受控的雾化水平，场景对象与摄像机的距离、雾化水平及地面实况（清晰图像）均已记录。数据集通过在封闭环境中使用专业雾化机控制雾化速度来创建，适用于评估不同去雾方法和模型的性能，特别是在计算机视觉问题如对象检测、深度估计等领域的应用。IMFD is a multi-level foggy image and video dataset developed by the Department of Computer Engineering at TED University, which aims to provide benchmark tests for defogging methods. This dataset contains two sets of color video data from real-world scenes with controlled fog levels. The distances between scene objects and the camera, the fog levels, and the ground truth (clear images) are all recorded. The dataset is constructed using professional fog generators in a closed environment to control the fogging speed, and it is suitable for evaluating the performance of various defogging methods and models, especially for applications in computer vision tasks such as object detection and depth estimation.

搜集汇总

背景与挑战

背景概述

IMFD是一个由泰德大学计算机工程系创建的室内多级雾化图像和视频数据集，专为去雾方法提供基准测试。它包含两个真实场景的彩色视频数据，具有受控的雾化水平和记录的距离、雾化水平及地面实况（清晰图像），通过在封闭环境中使用专业雾化机创建，适用于评估去雾方法在计算机视觉任务如对象检测和深度估计中的性能。