The invention provides a blind-restoration method for protecting motion blur images and relates to the field of image processing. The method adopts a multiscale strategy from coarseness to fineness: firstly, an image pyramid is established according to observed motion blur images; secondly, the following three steps are iteratively executed in each resolution layer of the image pyramid: 1, a first-order difference operator is applied to obtain space scale information of image edges, then Lo sparsity constraints of image gradient are combined to propose an image large-scale edge extracting model; 2, a dual regularization constraint item combining with Gaussian prior and Lo sparsity constraints is proposed to derive an estimation model of a motion blur degeneration function, and accurate estimation of the motion blur degeneration function is achieved; 3, a cost function for image restoration is derived to achieve restoration of clear and sharpened images; 4, a semi-secondary dissociatedinteractive optimized strategy is adopted to conduct optimized calculation on the proposed model, and the effectiveness of the method is proved through theoretical analysis and simulation results.