Image Deblurring with Impulse Noise Using Split Bregman Algorithm

We propose an effective method to resolve blurred images with impulse noise. Our method has two steps. First, an improved adaptive median filter is proposed for image denoising; And second, the problem of deblurring the denoised image is formulated as to minimize the object function which consists of L1 data-fedility term and double regularization term. The minimization problem is solved by split Bregman algorithm. Numerical results using image with different blurs and impulse noise show that the proposed method gives better performance than the variable splitting alternative minimization algorithm in [10] by objective peak signal to noise ratio and subjective vision quality, which demonstrates the efficiency of our proposed algorithms.