提出一种以高光谱图像分析为目标的基于二维主成分分析的高光谱遥感图像的降维方法.通过多变量线性变换对高光谱数据进行特征提取,应用二维主成分分析的方法对高光谱遥感图像进行降维.对AVIRIS图像应用二维主成分分析的方法,可将能量主要集中在少数几个特征值中,这就为降维提供了可能.计算机仿真结果表明,该算法计算量小,方差小,峰值信噪比(PSNR)、分类准确性均显著提高,MSE有所下降.%Based on Two Dimensional Principal Analysis (TDPCA) for hyperspectral image analysis, a dimensionality reduction method of hyperspectral images is introduced. The hyperspectral features are extracted by multilinear transformation and the dimensionality reduction is carried out by Two Dimensional Principal Analysis. Since the energy of the AVIRIS images mainly centralizes in a few eigenvalues, it is feasible for hyperspectral images dimensionality reduction. The experimental results show that the method achieved little computation, lower variance, greatly improved PSNR and classification accuracy with slight decline of MSE.
展开▼