MSD

MSD is a infrared and visible image dataset collected in real-world road environments. It comprises aligned infrared and visible images captured in multiple scenarios, such as dense fog, dusk, nighttime, intersection, and tunnel. While most existing public infrared and visible image datasets lack sufficient coverage of naturally occurring adverse conditions such as dense fog and dusk, MSD provides multi-modal images captured under real-world dense fog, low-light at dusk, and other scenarios, offering a valuable data option for developing and evaluating image fusion methods in diverse and realistic scenarios. MSD contains 208 visible images and 208 corresponding infrared images, and the infrared and visible images can be utilized to test the performance of infrared and visible image fusion methods or as part of a training set.MSD是一个采集自真实道路环境的红外与可见光图像数据集。该数据集包含多场景下采集的对齐后的红外与可见光图像，涵盖浓雾、黄昏、夜间、交叉路口及隧道等多种场景。当前多数公开的红外与可见光图像数据集，未能充分覆盖浓雾、黄昏这类自然形成的恶劣成像条件，而MSD提供了真实场景下的多模态图像（multi-modal images），包括真实浓雾、黄昏低光照环境等场景，为在多样且贴合实际的场景中开发与评估图像融合方法提供了极具价值的数据选项。MSD共包含208幅可见光图像与208幅对应的红外图像，此类图像既可用于测试红外与可见光图像融合方法的性能，也可作为训练集的组成部分。

搜集汇总

背景与挑战

背景概述

MSD是一个真实道路环境中的红外和可见光图像数据集，包含多种场景如浓雾、黄昏和夜间，弥补了现有数据集对自然恶劣条件覆盖不足的问题。它提供208对对齐的红外和可见光图像，可用于开发和评估图像融合方法，适用于多样化的现实场景。