In this paper, we present two algorithms to solve some inverse problems coming from the field of image processing. The problems we study are convex and can be expressed simply as sums of lp-norms of affine transforms of the image. We propose 2 different techniques. They are - to the best of our knowledge - new in the domain of image processing and one of them is new in the domain of mathematical programming. Both methods converge to the set of minimizers. Additionally, we show that they converge at least as O(1/N) (where N is the iteration counter) which is in some sense an ``optimal'' rate of convergence. Finally, we compare these approaches to some others on a toy problem of image super-resolution with impulse noise.