FMB Dataset

该数据集名为FMB，包含了1500对红外与可见光图像，涵盖了包括道路、人行道、建筑物和摩托车在内的14个类别，适用于多种环境条件。此外，该数据集还覆盖了雾、大雨和低光照等多样化条件，规模达到了1500张图像，其任务是进行语义分割。

This dataset, named FMB, contains 1500 pairs of infrared and visible-light images, covering 14 categories including roads, sidewalks, buildings, motorcycles and other classes. It is applicable to various environmental conditions, and additionally covers diverse scenarios such as fog, heavy rain and low-light environments, with a total of 1500 images. The task of this dataset is semantic segmentation.

搜集汇总

背景与挑战

背景概述

FMB Dataset是一个用于图像融合和分割的全时多模态基准测试数据集，属于SegMiF项目的一部分，可通过Google Drive和Baidu Yun下载。