基于加权仿射变换算法的高光谱数据降维

Hyperspectral Data Dimension Reduction Based on Weighted Robust Affine Set Fitting Algorithm

提出一种基于噪声白化和端元提取的加权仿射变换算法用于高光谱图像数据降维,相比较于基于端元提取的仿射变换算法,通过该算法降维后数据的信噪比更高,同时对原始信息的保存量更大,波段之间的相关性更低,从而表明了该算法的有效性.

A robust affine set fitting algorithm based on noise whitening and endmember extraction for hyperspectral image data dimension reduction is presented in this paper. Compared to the effect of robust affine set fitting algorithm based on endmember extraction, the data received by this method has a higher signal-to-noise ratio, and the original information is retained larger, at the same time the correlation of adjacent bands is lower, so as to verify the effectiveness of the proposed algorithm.