Subspace-based hyperspectral image band selection method

The invention provides a hyperspectral image band selection method based on a subspace, and relates to a hyperspectral band selection method. According to the method, a maximum ellipsoid volume wave band selection algorithm and a sequential forward algorithm are combined, wave band selection is carried out on hyperspectral data, meanwhile, the matrix covariance and inversion operation complexity in the operation process are reduced through the characteristics of matrix subspace and the iteration thought, the calculation time is shortened, and the calculation efficiency is improved. Therefore, the optimization of the wave band selection algorithm is achieved. 本发明提出一种基于子空间的高光谱图像波段选择方法，涉及一种高光谱波段选择方法。该方法将最大椭球体积波段选择算法与序贯前向算法结合起来，对高光谱数据进行波段选择，同时利用矩阵子空间的特点以及迭代思想来降低运算过程中矩阵协方差以及求逆的运算复杂度，缩短计算时间，从而达到对波段选择算法的优化。