"PanCollection" for Remote Sensing Pansharpening

[GitHub Page]

 

本页面主要分享当前基于深度学习的遥感图像全色锐化的一些资源，如下： 1）发布相关卫星（如WorldView-3, QuickBird, GaoFen2，WorldView-2卫星）的全色锐化训练-测试数据集PanCollection； 2）发布基于Pytorch深度学习库的Python代码统一编写框架，便于后来学者的入门、开展研究以及公平比较；读者可以直接将上述数据集放到此代码框架，快速进行训练和测试得到多个深度学习方法的结果。 3）发布统一的全色锐化传统-深度学习方法MATLAB测试软件包，便于后来学者进行公平的测试。

(1) PanCollection Dataset (PanCollection 数据集) [GitHub Page] WV3 Dataset [1] Training Dataset(训练数据集, 5.76GB)： [Baidu Cloud] [2] Testing Dataset(测试数据集, 20 Examples/per class)： [ReducedData(H5 Format)] [FullData(H5 Format)] Or, you can use the equivalent seperated .mat examples:： [ReducedData(Mat Format)] [FullData(Mat Format)] QB Dataset [1] Training Dataset(训练数据集, 5.37GB)： [Baidu Cloud] [2] Testing Dataset(测试数据集, 20 Examples/per class)： [ReducedData(H5 Format)] [FullData(H5 Format)] Or, you can use the equivalent seperated .mat examples： [ReducedData(Mat Format)] [FullData(Mat Format)] GF2 Dataset [1] Training Dataset(训练数据集, 6.21GB)： [Baidu Cloud] [2] Testing Dataset(测试数据集, 20 Examples/per class)： [ReducedData(H5 Format)] [FullData(H5 Format)] Or, you can use the equivalent seperated .mat examples： [ReducedData(Mat Format)] [FullData(Mat Format)] WV2 Dataset [1] Testing Dataset(测试数据集, 20 Examples/per class)： [ReducedData(H5 Format)] [FullData(H5 Format)] Or, you can use the equvilent seperated .mat examples： [ReducedData(Mat Format)] [FullData(Mat Format)]

(2) Uniformed Pytorch Coding Framework (统一的Pytorch代码框架) [1] Download Link(下载链接)： [GitHub], or similar toolbox [DLPan-Toolbox] [2] One example to illustrate details(一个简单示例，推荐通过此学习)： [preparing]

(3) Matlab Testing Package for Trad. & DL Methods (传统方法-深度学习方法测试MATLAB软件包) [1] Download Link： [DLPan-Toolbox]/02-Test-toolbox-for-traditional-and-DL(Matlab) [2] One example to illustrate details(一个简单使用示例，推荐通过此学习)： [preparing]