A REVIEW PAPER ON VIDEO RESTORATION VICTIMIZATION NON-LOCAL KERNEL REGRESSION

Author’s Name : Kavitha M

Volume 01 Issue 01  Year 2014  ISSN No:  2349-3828  Page no: 26-27

Abstract -This paper presents a non-local kernel regression (NL-KR) model is figure for numerous video restoration tasks. the strategy exploits each the non-local self-similarity and native structural regularity properties in natural pictures. The non-local self-similarity is predicated on the observation that image patches tend to repeat themselves in natural pictures and videos; and also the native structural regularity observes that image patches have regular structures wherever correct estimation of element values via regression is feasible. during this work, we tend to apply the planned model to video de-noising, de-blurring and super-resolution reconstruction. intensive experimental results on each single pictures and realistic video sequences demonstrate that the planned framework performs favourably with previous works each qualitatively and quantitatively.

Keywords –Local structural regression, non-local self similarity, restoration, denoising, deblurring, super-resolution