Hyperspectral Imaging Dataset

Dataset Description

This repository contains hyperspectral imaging (HSI) datasets and their corresponding ground-truth data used for research and experimental evaluation of hyperspectral image analysis and classification methods. The dataset files are provided in MATLAB (.mat) format and include the hyperspectral image data together with the corresponding ground-truth labels.

The HSI data consist of spatially organized spectral measurements, where each pixel is represented by a high-dimensional spectral signature. The associated ground-truth files provide the reference class labels required for supervised learning, classification, quantitative evaluation, and comparison of HSI processing algorithms.

The repository is intended to facilitate reproducible research in hyperspectral image processing, spectral-spatial analysis, classification, and deep learning. Researchers can use the provided HSI data and ground-truth labels for algorithm development, model training and testing, benchmarking, and comparative evaluation.

Contents

• HSI data: Hyperspectral image cubes provided in .mat format.

• Ground-truth data: Corresponding reference label maps provided in filename_gt.mat format.

• Metadata: Information describing the datasets, data organization, and associated files, where applicable.

Users are requested to cite the associated publication(s) and this Zenodo dataset when using these data in academic or research work.