A nonlocal feature-driven exemplar-based approach for image inpainting

TL;DR

This paper introduces a nonlocal variational method for image inpainting that effectively restores multiple structures and textures by leveraging self-similarity and a novel anisotropic patch distance metric.

Contribution

It proposes a unified framework combining convolution-based geometric recovery with exemplar-based texture synthesis, including a new anisotropic patch distance metric and an optimization algorithm.

Findings

Effective inpainting of multiple structures and textures

Improved control over feature selection with anisotropic metric

Validated through experiments on various test images

Abstract

We present a nonlocal variational image completion technique which admits simultaneous inpainting of multiple structures and textures in a unified framework. The recovery of geometric structures is achieved by using general convolution operators as a measure of behavior within an image. These are combined with a nonlocal exemplar-based approach to exploit the self-similarity of an image in the selected feature domains and to ensure the inpainting of textures. We also introduce an anisotropic patch distance metric to allow for better control of the feature selection within an image and present a nonlocal energy functional based on this metric. Finally, we derive an optimization algorithm for the proposed variational model and examine its validity experimentally with various test images.