高光谱遥感图像的稀疏分解与压缩感知

Compressed Sensing and Sparse Decomposition of Hyperspectral Remote Sensing Images

基于稀疏分解和压缩感知原理对高光谱遥感图像进行压缩重构,提出一种基于过完备原子库上分解图像的稀疏分解快速算法,以减少图像稀疏分解的计算量.仿真计算结果表明,利用压缩感知和谱带分组技术对高光谱图像进行谱间压缩,可提高运算速度,并降低成本.

In order to improve sampling of the compression coding technology and constantly improve the imaging spectrometer systems,based on the principles of the sparse decomposition and compressed sensing,hyper-spectral remote sensing images were compressed and reconstructed.A fast sparse decomposition algorithm of the image decomposition on over-complete atomic library was proposed so as to reduce the amount of computing.The simulation results show that the calculation improved processing speed and reduced costs by spectral compression of hyper-spectral images carried out with the compressed sensing and bands grouping technology.