Quantifying Extraneous Matter in Shredded Sugarcane Using Near-Infrared Spectroscopy

Abstract

Sugar cane is one of the most important agricultural commodities globally, serving as a vital source of sugar, bioethanol, and employment for millions of people across more than 100 countries. Rising rainfall due to climate change, evolving environmental regulations, and cost-driven harvesting practices have increased extraneous matter (EM) in sugar cane, reducing factory efficiency and sugar recovery. Despite its significant impact, EM cannot be regularly quantified in sugar factories due to the lack of practical measurement methods and is therefore still excluded from cane payment systems. We introduce near-infrared (NIR) spectroscopy as a rapid, nondestructive solution to this challenge. NIR calibration models for leaf content in shredded cane were developed using mixtures of clean cane, soil, and leaves with known EM concentrations, and soil content calibrations were built using incinerated ash as the reference method. Partial least-squares regression models with k-fold cross-validation were developed to correlate NIR spectra with reference values. Soil content based on ash analysis yielded strong calibration results (R² = 0.88), markedly outperforming sediment analysis (R² = 0.12). For the first time, NIR successfully predicted brown leaves (R² = 0.72), green leaves (R² = 0.73), and total leaves (R² = 0.88). These findings prove the potential of NIR spectroscopy to revolutionize EM analysis, providing a practical pathway for its integration into cane payment systems and improving sugar cane quality assessment.