The nonlinear method of suppression of an additive noise on digital image based on the preliminary image processing using the surrogate data technology and singular value decomposition is proposed in the paper. It allows to save the resolution capability in the case of higher noise level as compared to linear window filtering method. The simulation results of preliminary processing of the image approximated with using of the image singular value decomposition and ATS (attractor trajectory surrogates) algorithm of surrogate image generation are presented. The comparative analysis of the resolution capability of proposed method with the method of image median filtering is performed.
surrogate images, ATS algorithm, image singular value decomposition