An Automated Algorithm of Peak Recognition Based on Continuous Wavelet Transformation and Local Signal-to-Noise Ratio

Abstract

Raman peaks carry valuable information about constituent chemical bonds. Therefore, peak recognition is a very essential part of spectral analysis. The fully automated peak recognition is convenient in practical application. A fully automated Raman peaks recognition algorithm based on continuous wavelet transformation and local signal-to-noise ratio (LSNR) is proposed. This algorithm extracts feature points through continuous wavelet transformation and recognizes peaks through LSNR. This algorithm also can be used to eliminate spike, noise, and baseline. Both simulated and experimental data are used to evaluate the performance of the CWT-LSNR algorithm compared with the other two algorithms. The results show that CWT-LSNR gives better accuracy and has the advantage of easy use.