A two-stage denoising filter: the preprocessed Yaroslavsky filter

08/31/2012
by   Joseph Salmon, et al.
0

This paper describes a simple image noise removal method which combines a preprocessing step with the Yaroslavsky filter for strong numerical, visual, and theoretical performance on a broad class of images. The framework developed is a two-stage approach. In the first stage the image is filtered with a classical denoising method (e.g., wavelet or curvelet thresholding). In the second stage a modification of the Yaroslavsky filter is performed on the original noisy image, where the weights of the filters are governed by pixel similarities in the denoised image from the first stage. Similar prefiltering ideas have proved effective previously in the literature, and this paper provides theoretical guarantees and important insight into why prefiltering can be effective. Empirically, this simple approach achieves very good performance for cartoon images, and can be computed much more quickly than current patch-based denoising algorithms.

READ FULL TEXT
research
05/01/2015

Image Denoising using Optimally Weighted Bilateral Filters: A Sure and Fast Approach

The bilateral filter is known to be quite effective in denoising images ...
research
02/22/2021

Shadow Image Enlargement Distortion Removal

This project aims to adopt preprocessing operations to get less distorti...
research
12/03/2012

An Image Based Technique for Enhancement of Underwater Images

The underwater images usually suffers from non-uniform lighting, low con...
research
11/12/2019

MRI denoising using Deep Learning and Non-local averaging

This paper proposes a novel method for automatic MRI denoising that expl...
research
10/21/2016

Multispectral image denoising with optimized vector non-local mean filter

Nowadays, many applications rely on images of high quality to ensure goo...
research
11/24/2011

Suboptimality of Nonlocal Means for Images with Sharp Edges

We conduct an asymptotic risk analysis of the nonlocal means image denoi...
research
03/01/2022

When A Conventional Filter Meets Deep Learning: Basis Composition Learning on Image Filters

Image filters are fast, lightweight and effective, which make these conv...

Please sign up or login with your details

Forgot password? Click here to reset