Staircasing reduction model applied to total variation based image reconstruction

Total variation based image restoration method was first proposed by Rudin Osher and Fatermi in 1992. The images resulting from its application are usually piecewise constant, and have sometimes undesirable staircasing effect. To reduce this effect, we propose an improved model by combining the advantages of total variation and H¹ regularization. The new model substantially reduces the staircase effect, while preserving sharp edges. This model can be used in image reconstruction, it has advantages of keeping edges and recovering smooth region´s value. We give 1D and 2D experimental results to show the efficiency of the proposed model.