Windowed total variation denoising and noise variance monitoring

TL;DR

This paper introduces a real-time Total-Variation denoising method with automatic hyper-parameter selection, adapted for non-stationary signals using sliding windows, and includes a noise variance monitoring approach validated by simulations.

Contribution

It develops a real-time TV denoising technique with automatic hyper-parameter tuning and extends it to non-stationary signals with noise variance monitoring.

Findings

The method performs well on simulated data.

It accurately tracks noise variance changes.

The approach is suitable for real-time applications.

Abstract

We proposed a real time Total-Variation denosing method with an automatic choice of hyper-parameter $\lambda$, and the good performance of this method provides a large application field. In this article, we adapt the developed method to the non stationary signal in using the sliding window, and propose a noise variance monitoring method. The simulated results show that our proposition follows well the variation of noise variance.