Gaussian noise parameter estimation based on multiple singular value decomposition and non?linear fitting

Noise standard deviation (STD) is an important parameter in many digital image processing applications. This paper presents a Gaussian noise estimation algorithm using multiple singular value decomposition (SVD) and non-linear fitting. The proposed adds known to the original times generate noise-corrupted set then performs SVD on each image. By analyzing values of images, overdetermined equation system with respect STD established. Gauss–Newton iteration method backtracking Armijo line search are used solve equations, which improve convergence speed reduce computational cost. Compared other methods, mean error TID2008 dataset 0.028, several lower than methods. shows that performance our estimator significantly improved.