Paper

Local Entropy Based Image Reconstruction


Authors:
Carsten Croonenbroeck
Abstract
This article presents a local entropy based image reconstruction algorithm that performs quite well in cases where there is distortion in an image. If the “wanted” image information is still available but distributed over two or more distinct images, the algorithm can collect the required information from the set of the images given. Instead of inferring pixel data information from the remainder of one single image, the algorithm provides a decision rule on what information from which one of the set of given images to actually use in order to create a new, (ideally) distortion-free image.
Keywords
Image Processing; Information Reconstruction; MSE; PSNR
StartPage
137
EndPage
141
Doi
Download | Back to Issue| Archive