High-resolution image via Multiselective Contourlets

This paper develops an interpolation algorithm for generating a high-resolution image from a low-resolution input image. The idea is to use a simple wavelet-based linear interpolation scheme as our initial high-resolution image; and an iterative projection process based on hard-thresholding scheme in a new Multiselective Contourlets domain, to improve the regularity of object boundaries in the generated images. Experimental results show that the new proposed algorithm achieves better high-resolution results than bicubic interpolation, and classic contourlet transform high-resolution methods, in subjective quality, and in most cases, in terms of PSNR as well.