Joint Baseline-Correction and Denoising for Raman Spectra

Abstract

Laser instruments often suffer from the problem of baseline drift and random noise, which greatly degrade spectral quality. In this article, we propose a variation model that combines baseline correction and denoising. First, to guide the baseline estimation, morphological operations are adopted to extract the characteristics of the degraded spectrum. Second, to suppress noise in both the spectrum and baseline, Tikhonov regularization is introduced. Moreover, we describe an efficient optimization scheme that alternates between the latent spectrum estimation and the baseline correction until convergence. The major novel aspect of the proposed algorithms is the estimation of a smooth spectrum and removal of the baseline simultaneously. Results of a comparison with state-of-the-art methods demonstrate that the proposed method outperforms them in both qualitative and quantitative assessments.

Blind Spectral Signal Deconvolution with Sparsity Regularization: An Iteratively Reweighted Least-Squares Solution

“…The parameters for the FSD, HOS, and SDTR methods were the same as in the original references ( [13,29], and [22], respectively) because the MATLAB source codes were available. For an overall evaluation, the following quantitative indices were used: the RMSE of the distortion of the peak position; noise suppression (NS), where NS = |∇g|/|∇x|; and width reduction (WR), where…”

“…Spectral signals have become the preferred tools for characterization and analysis of materials [2,12,13,18,23,25], but they often are affected by band overlap and random noise, which limit their use. The spectral signal as measured by a dispersion spectrophotometer can be mathematically modeled as a latent spectrum convoluted with the instrument response function (IRF), i.e.,…”

“…The major factors part of the IRF that limit the resolving power are the slit width and diffractive limit of the dispersive grating [4]. Different approaches to spectral deconvolution have been proposed over the recent past [1,6,13,14,16,17,19,22,26,27,29]. The existing algorithms can be classified into two major categories: curve-fitting [1,16] and spectral deconvolution [6,22,27,29].…”

A practical convolutional neural network model for discriminating Raman spectra of human and animal blood

“…Liu et al proposed an iterative baseline correction algorithm named Goldindec, which not only suppress the influence of peaks but also solves the problem of low‐correction accuracy when there is a high‐peak number. Liu et al presented a model to estimate a smooth spectrum and remove the baseline simultaneously. Fan et al recovered the Raman spectra from severe background noise based on sparse representation trained by Batch‐Orthogonal Matching Pursuit (Batch‐OMP).…”

Salient space detection algorithm for signal extraction from contaminated and distorted spectrum

“…18 Algorithms that do not use a model of the baseline or signal shape and that have anti-noise capability are preferred. Many iterative algorithms [19][20][21] and Difference-of-Gaussian (DoG) functions can meet this requirement. Adaptive iteratively reweighted penalized least squares (airPLS) is a wellknown iterative algorithm, because it is flexible and valid; however, it needs further optimization.…”