Edge Preserving Single Image Super Resolution Techniques– A Comprehensive Study

High-Resolution (HR) images play vital role in almost every aspect of day-to-day life. The spatial resolution and the quality of the images can be improved with help of Super Resolution (SR) techniques. It rebuilds a HR image from one or multiple LowResolution (LR) images. During the application of these Super Resolution (SR) methods, some intricate details in the given low resolution image may be lost. Minute details preservation is essential in areas like medical imaging. This paper makes an attempt to review some of the state of the art single image edge preservation super resolution techniques. The current comprehensive survey classifies SR methods broadly into interpolation based, reconstruction based and learning based methods. Interpolation based methods are straight forward, but ends with producing blurred details and ringing artefacts. Reconstruction based methods are fast, but suffer with optimized parameters setting which are required for fine tuning of the output. Learning based methods involves construction of large image dictionaries. Though training these dictionaries consumes more time and complicated, learning based methods yield better quality images. Electing and designing a proper SR technique can be a real challenge for the researcher. However, assimilation of these methods are always proved to elicit better result.