The Single image super resolution using Lifting-Based Directionlets

Main Article Content

Reji A P

Abstract

This paper proposes, a fast learning based single image super resolution method which can be used for real time
applications using multiple direction wavelet transform, called Directoionlets. Conventional directionlet transform is computationally very intensive, so a lifting based implementation is used
here for super resolving both grey and colour images. Here two methods, directional variance
method and skewed wavelet transform are used in implementing directionlet transform. From
the results it is clear that directional variance method is faster than skewed wavelet method.
Results using different wavelets like db4 ,bior3.3 etc are compared here. The simulation results
showed that the proposed approach is faster and needs less memory compared to conventional
directionlet based single image super resolution.

Article Details

How to Cite
A P, R. (2021). The Single image super resolution using Lifting-Based Directionlets. INFOCOMP Journal of Computer Science, 20(2). Retrieved from https://infocomp.dcc.ufla.br/index.php/infocomp/article/view/1767
Section
Computer Graphics, Image Processing, Visualization and Virtual Reality