Locally adaptive image denoising by a statistical multiresolution criterion

TL;DR

This paper introduces a method for locally adaptive image denoising using a statistical multiresolution criterion to select smoothing parameters, effectively detecting edges and improving denoising quality in applications like microscopy.

Contribution

It proposes a novel approach for local parameter selection in image denoising, combining statistical multiresolution criteria with inhomogeneous diffusion and total variation regularization.

Findings

Effective local denoising demonstrated on numerical examples

Smoothing parameter acts as an edge detector

Application shown in confocal microscopy

Abstract

We demonstrate how one can choose the smoothing parameter in image denoising by a statistical multiresolution criterion, both globally and locally. Using inhomogeneous diffusion and total variation regularization as examples for localized regularization schemes, we present an efficient method for locally adaptive image denoising. As expected, the smoothing parameter serves as an edge detector in this framework. Numerical examples illustrate the usefulness of our approach. We also present an application in confocal microscopy.