Wavelet analysis of near infrared spectral data in the application of denoising

摘要

Spectrum signal may contain many peaks or mutations and noise also is not smooth white noise, to this kind of signal analysis, must do signal pretreatment, remove part of signal and extract useful part of signal. Based on the data of blood glucose near-infrared spectrum as the research object to explore the application of wavelet transform in the near infrared spectrum signal denoising, and through the simulation results show that using wavelet analysis of near infrared spectral data pretreatment than the traditional Fourier method can be higher precision of prediction.