Retrieval of nicotine content in cigar leaves by remote analysis of aerial hyperspectral combining machine learning methods
January 31, 2025 by nature.com
Retrieval of nicotine content in cigar leaves by remote analysis of aerial hyperspectral combining machine learning methods
The experiment included testing 15 cigar varieties with a split-plot design, focusing on nitrogen fertilizer levels. Spectral data from tobacco leaves were collected using a UAV equipped with a hyperspectral camera. Signal processing techniques were employed to enhance spectral features, and variable selection methods like UVE, SPA, and CARS were used. Inversion models such as PLSR, BP, SVR, and RFR were applied with R2 and RMSE as evaluation metrics for predicting nicotine content. A 5-fold cross-validation approach and RPD were utilized to assess model performance. Matlab 2021a was used for all data analysis and evaluations.
