Signal denoising based on the Schr\"odinger operator's eigenspectrum and a curvature constraint
Abstract
Recently, a new Signal processing method, named Semi-Classical Signal Analysis (SCSA), has been proposed for denoising Magnetic Resonance Spectroscopy (MRS) signals. It is based on the Schr\"odinger Operator's eigenspectrum. It allows an efficient noise reduction while preserving MRS signal's peaks. In this paper, we propose to extend this approach to different signals, in particular pulse shaped signals, by including an optimization that considers curvature constraints. The performance of the method is measured by analyzing noisy signal data and comparing with other denoising methods. Results indicate that the proposed method not only produces good denoising performance but also guarantees the peaks are well preserved in the denoising process.
Keywords
Cite
@article{arxiv.1908.07758,
title = {Signal denoising based on the Schr\"odinger operator's eigenspectrum and a curvature constraint},
author = {Peihao Li and Taous Meriem Laleg-Kirati},
journal= {arXiv preprint arXiv:1908.07758},
year = {2019}
}