Here we provide the resources to reproduce the results presented in the following article
Rasti, B., Ghamisi, P., "Hyperspectral Denoising: From Conventional Techniques Towards Deep Learning Ones", 1st workshop on Complex Data Challenges in Earth Observation (CDCEO)1 Nov. 2021.
The following HSI denoising methods: 3D Wavelets,FORPDN, and HyRes all can be found in Hyperspectral Image Denoising Matlab Toolbox, SSTV, NAIRLMA, FastHyDe, DIP, For DIP, you need to download the original toolbox and copy the python file (DN4Rev_Real.py) into the master folder (change the path, line 48 of DN4Rev_Real.py, to the mat file of Indian Pine) and run. Addtionally, you need to install the dependencies (see readme)
SDeCNN. For SDeCNN, after downloading the codes only replace the Indian.mat which has 206 bands with the one that we used with 220 bands (The link to the dataset is given below).
To reproduce the resutls shown below you can find the Indian Pines dataset in Hyperspectral Image Denoising Matlab Toolbox and use the matlab and python demos provided to run the codes
