基于冗余字典的高光谱图像的稀疏分解

摘要

基于冗余字典的稀疏表示方式能够以较少的数据量，更好地描述高光谱图像中的特征信息，是一种更有效的高光谱图像表示方法。根据高光谱图像自身的特点，使用梯度下降法学习冗余字典，首先固定字典，用梯度下降法训练系数；然后固定系数，再用同样的方法训练字典，以上两步交替进行，直到算法收敛；最后将这个字典用于高光谱图像的重构。实验结果表明，该方法获得了良好的重构效果。%A sparse representation based on redundant dic-tionary can use a small amount of data to describe the fea-ture of hyperspectral image better. It is a more effective method to represent hyperspectral image. According to its own characteristics,a gradient descent method can be used to learning redundant dictionary. First,assuming the dictionary is fixed,gradient descent method is used to calculate the sparse coefficients;second,assuming the coefficients are fixed,in the same way we can update the dictionary,these two steps are alternately used until the algorithm converges. Finally,the dictionary is applied to reconstruct hyperspectral images. Experimental results show that the method can ob-tain a good reconstruction results.