A double constrained soft-decision image interpolation algorithm

The key to image interpolation is recovering pristine edge and detailed information as much as possible. Robust soft-decision adaptive interpolation (RSAI) shows excellent interpolation performance by adding weighted least squares estimation to soft-decision adaptive interpolation (SAI).We put forward a novel interpolation strategy-double constraints to modify and improve RSAI. Every pixel is double bound during data estimation, by which edge details and textures are better preserved. Experimental results demonstrate that our proposed algorithm produces the best interpolation results among state-of-the-art algorithms based on PSNR and SSIM objective evaluations.