Auto-grouping method of flue-cured tobacco leaves based on near infrared spectra technology

In order to explore the nondestructive testing technology of flue-cured tobacco leaf on purchasing quality, the new method of fast classify tobacco leaf’s group (parts, color) was proposed based on near infrared spectral technology. The feasibility of applying near infrared spectral technology to evaluate the complete flue-cured tobacco leaves quality was analyzed.. The prediction performances of different band range, different principal component numbers and different preprocessing methods of the spectra (multiplicative signal correction, standard normal variation, and derivative spectra) together with discriminant analysis (DA) were also investigated, and the calibration model was respectively established to classify the different parts and color of the flue-cured tobacco leaves. The research results for the tobacco leaf classification showed that DA calibration model using the parameters of band range between 1?101 and 2?395 nm combined with original spectra was optimal, which the correct percentage of classification on part and color of tobacco leaf was 100% for calibration sets, and it was 98.6% and 97.1% respectively for validation set. It is proved that the new method proposed in this study is capable to discriminate the parts and color of tobacco leaves with high accuracy. In addition, it might provide a new method to discriminate tobacco leaves group.