Efficient Upsampling of Natural Images

TL;DR

This paper introduces an efficient image upsampling method that separately models and enhances edges and textures, producing high-resolution images with sharp details and minimal artifacts without requiring training.

Contribution

It presents a novel approach that decomposes images into edge and detail layers, then separately synthesizes and merges them using a non-convex energy minimization framework, avoiding training phases.

Findings

Accurately reproduces sharp edges and textures.

Avoids common artifacts like ringing and haloing.

No training or parameter estimation needed.

Abstract

We propose a novel method of efficient upsampling of a single natural image. Current methods for image upsampling tend to produce high-resolution images with either blurry salient edges, or loss of fine textural detail, or spurious noise artifacts. In our method, we mitigate these effects by modeling the input image as a sum of edge and detail layers, operating upon these layers separately, and merging the upscaled results in an automatic fashion. We formulate the upsampled output image as the solution to a non-convex energy minimization problem, and propose an algorithm to obtain a tractable approximate solution. Our algorithm comprises two main stages. 1) For the edge layer, we use a nonparametric approach by constructing a dictionary of patches from a given image, and synthesize edge regions in a higher-resolution version of the image. 2) For the detail layer, we use a global parametric texture enhancement approach to synthesize detail regions across the image. We demonstrate that our method is able to accurately reproduce sharp edges as well as synthesize photorealistic textures, while avoiding common artifacts such as ringing and haloing. In addition, our method involves no training phase or estimation of model parameters, and is easily parallelizable. We demonstrate the utility of our method on a number of challenging standard test photos.