Raman hyperspectral image analysis of benzoyl peroxide additive

This study adopted a Raman hyperspectral imaging system to collect the Raman spectra and hyperspectral images of benzoyl peroxide (BPO) additive for subsequent analysis. The raw Raman spectra of BPO were preprocessed by wavelet denoising. Optimal parameters of wavelet denoising were selected by the orthogonal experimental design. The signal-noise (S/N) ratio of the optimal parameter combination was 32.848. The smoothed Raman spectra were divided into three regions (1900–1300, 1300–700, and 700–100 cm−1) for assignment and the band vibrational modes of BPO molecule were obtained. Wherein, the Raman bands at 1771, 1597, 1230, 999, 889 and 845 cm−1 are higher than others, and can be used as the Raman characteristic bands. Analysis of the grayscale images corresponding to different characteristic bands, it is found that the order of change of the image gray level was consistent with characteristic bands intensity. Findings of this study provide the research basics for the detection and analysis of BPO additive.