Super resolution image by edge-constrained curve fitting in the threshold decomposition domain

An edge-constrained curve-fitting method is proposed in this paper to produce super-resolution (SR) images from a low-resolution (LR) source image. The novelty of this method lies that the threshold decomposition is applied on the source image to obtain multiple binary images, and then an edge-constrained curve-fitting is applied on the resulting set of binary images. This allows us to focus on tiny objects and thin structures so as to achieve rather nice visual results even when a large zoom-in factor is used. Our results are compared with those achieved by using the bi-cubic interpolation, showing the ability of our algorithm to achieve much better visual quality in smooth areas as well as for sharp edges and small objects.