Detection of tobacco blend ratio based on hyperspectral imaging

YANG Tianzhuo; HE Jin; WU Lianlian; DAI Yongsheng; YI Bin; LI Huajie; ZHANG Erqiang; DU Jinsong

doi:10.12187/2025.03.013

Abstract：This study focuses on detecting tobacco blend ratios using hyperspectral imaging.Due to the lack of rapid methods for detecting tobacco blend ratios on cigarette production lines,spectral data were collected from mixed tobacco with different blend ratios using hyperspectral imaging technology and machine learning methods.The effects of single and combined preprocessing techniques on model performance were explored.Regression models were established using partial least squares regression(PLSR)and support vector machine regression(SVR).Feature wavelength selection was performed with least angle regression(LARS),successive projections algorithm(SPA),competitive adaptive reweighted sampling(CARS),and genetic algorithm(GA)to build simplified models.The results showed that preprocessing methods,either individually or combined,affected model accuracy.The combined wavelet transform and SG filtering(Wave+SG)method reduced the mean absolute percentage error(MAPE)by 1.2％compared to raw spectral data.The Wave+SG-GA-PLSR model performed best,with MAPE values of 1.415％and 1.531％for the training and test sets in two-component blends,respectively.This method was also applicable to multi-component blends,with MAPE values below 8.361 5％for both three-component and four-component blends.Hyperspectral imaging combined with machine learning can accurately predict the proportions of components in mixed tobacco,providing a reference for online monitoring of blend uniformity and quality control in cigarette production.