实验室创建的用于高光谱解混评估的具有地面实况的数据集
收藏资源简介:
该数据集由西北工业大学海洋科学与技术学院创建,包含36种混合场景,模拟不同混合模型,每个像素包含256个光谱带,光谱分辨率高达0.58nm。数据集提供纯材料的光谱签名及其组成,支持对解混算法进行定量和客观的评估。应用领域包括高光谱图像处理,旨在解决像素可能包含多种材料的问题,通过解混将像素光谱分离成一组端元和相应的丰度。
This dataset was developed by the School of Marine Science and Technology, Northwestern Polytechnical University. It includes 36 mixed scenarios simulating various mixing models, with each pixel containing 256 spectral bands and a spectral resolution as high as 0.58 nm. The dataset provides spectral signatures and compositional information of pure materials, enabling quantitative and objective evaluations of spectral unmixing algorithms. Its application fields cover hyperspectral image processing, which aims to address the issue that a single pixel may contain multiple materials, and separates the pixel's spectrum into a set of endmembers and their corresponding abundances via spectral unmixing.
- 1A laboratory-created dataset with ground-truth for hyperspectral unmixing evaluation西北工业大学海洋科学与技术学院 · 2019年



