线性混合模型的光谱解混算法综述

A review on spectral unmixing algorithms based on linear mixing model

混合像元的存在不仅影响了基于高光谱图像的地物识别和分类精度,而且已经成为遥感科学向定量化发展的主要障碍。目前的混合像元分解算法大多采用线性混合模型,其关键步骤为端元提取。文中从线性混合模型的定义出发,总结了近年来提出的端元提取算法,并重点对SMACC、VCA、SGA等算法进行了深入的分析,最后总结了混合像元分解的发展趋势。

The mixels not only influence the accuracy of target detection and classification based on the hypersepectral images, but also greatly hinder the development of quantitative remote sensing. Now, most of the spectral unmixing algorithms are based on the linear spectral mixing model, and the model＇ s key procedure is endmember extraction. This paper firstly introduced the linear mixing model, then summarized the endmember extraction algorithms and deeply analyzed several endmember extraction algorithms （ SMACC, VCA, SGA, etc） . After that, the trends of the spectral unmixing were summarized.