Super Resolution of an Image using Projection onto Convex Set Algorithm

Recently single image super resolution is important research area to generate high-resolution images from given low-resolution images. The term Super resolution is a process to increase the resolution of an image. This improvement quality of image is due to sub-pixel shift of low-resolution images from each other between images. In this paper, we concentrate on the problem of producing a high-resolution image from a single or multiple low-resolution images. We propose a POCS algorithm for the canonical problem of super-resolution (SR) image, which will obtain from low-resolution images. Single image super resolution algorithms are mainly based on frequency domain and spatial domain. Here image super resolution algorithm is presented which based on both spatial and frequency domain and consider the advantage of both. Algorithm is flexible and use back projection to minimize reconstruction error. Compared with traditional algorithm, experiment results show that the POCS algorithm eliminates the halo effect, thus the reconstructed image can achieve a good visual effect.