Skip to Main Content
Raman spectral interpretation often suffers common problems of band overlapping and random noise. Spectral deconvolution and feature-parameter extraction are both classical problems, which are known to be difficult and have attracted major research efforts. This paper shows that the two problems are tightly coupled and can be successfully solved together. Mutual support of Raman spectral deconvolution and feature-extraction processes within a joint variational framework are theoretically motivated and validated by successful experimental results. The main idea is to recover latent spectrum and extract spectral feature parameters from slit-distorted Raman spectrum simultaneously. Moreover, a robust adaptive Tikhonov regularization function is suggested to distinguish the flat, noise, and points, which can suppress noise effectively as well as preserve details. To evaluate the performance of the proposed method, quantitative and qualitative analyses were carried out by visual inspection and quality indexes of the simulated and real Raman spectra.
Instrumentation and Measurement, IEEE Transactions on (Volume:62 , Issue: 2 )
Date of Publication: Feb. 2013