MuRA-T

MuRA-T数据集是由普渡大学创建的，用于变化检测研究的多时相多分辨率卫星图像数据集。该数据集通过增强SpaceNet-7数据集，添加了时间平行的Landsat和Sentinel图像数据堆栈，每个AOI包含约24个月的图像数据。数据集的创建过程中，使用了MuRA框架进行跨卫星图像的空间对齐，确保高分辨率和低分辨率图像的精确对齐。MuRA-T数据集特别适用于研究城市发展、自然灾害评估和土地覆盖变化等领域，旨在解决多卫星数据融合中的变化检测问题。

The MuRA-T dataset is a multi-temporal and multi-resolution satellite image dataset created by Purdue University for change detection research. This dataset is developed by augmenting the SpaceNet-7 dataset, with added temporally parallel Landsat and Sentinel image data stacks, and each AOI contains approximately 24 months of image data. During the dataset construction process, the MuRA framework was employed to perform spatial alignment across satellite images, ensuring precise alignment between high-resolution and low-resolution images. The MuRA-T dataset is particularly suitable for research in fields such as urban development, natural disaster assessment, and land cover change, aiming to address the challenge of change detection in multi-satellite data fusion.

搜集汇总

背景与挑战

背景概述

MuRA-T是一个用于卫星图像变化检测的数据集，扩展自SpaceNet 7挑战数据集，专注于多时相多分辨率图像的对齐。它提供Landsat和Sentinel数据，适用于计算机视觉和模式识别研究，由普渡大学团队于2023年发布，遵循GNU GPL v3许可证。