A new measurement matrix optimal algorithm based on SVD

This paper proposes a new compressed sensing (CS) measurement matrix optimal algorithms based on singular value decomposition (SVD). New measurement matrix can be obtained by using SVD for the decomposition of Gaussian measurement matrix. Simulation results prove that using the new measurement matrix can not only greatly improve the robustness and stability of CS algorithm, but also have better behaviors on image quality recovery. Moreover, this method is suitable for the further study of other random measurement matrix.