Estimation of Sugar Content in Sugarcane (Saccharum spp.) Variety Lumpang 92-11 (LK 92-11) and Khon Kaen 3 (KK 3) by Near Infrared Spectroscopy

Abstract

In this study, a non-destructive measuring method, near-Infrared (NIR) technique was used to evaluate the quality of sugarcane. Two sugarcane (Saccharum spp.) varieties viz., Lumpang 92-11 and Khon Kaen 3 were chosen for the test. The samples were collected for 3 years. The sugar contents were measured in terms of ^°Brix, %Pol, %Fiber, and Commercial Cane Sugar (CCS) values using the NIR technique and conventional laboratory testing for comparison. The Partial Least Squares Regression (PLSR) model was performed using 400 samples for each variety. The NIR models showed the coefficient of determination (R²) of 0.97, 0.90, 086 and 0.82 for ^°Brix, %Pol, %Fiber and CCS, respectively with the corresponding root mean square error of prediction (RMSEP) of 0.246, 0.512, 0.353 and 0.542. The results indicated that the modelling using ^°Brix gave the best estimation with the highest R² and lowest RMSEP, indicating high accuracy and reliability. The modelling with %Pol and %Fiber gave the moderate estimations and that with CCS value gave the lowest accuracy. However, all the four modelling predictions were within the acceptable range and could thus be used in crops trading instead of the traditional method. It was more reliable, quicker, more comfortable and more environmentally friendly than the traditional method as it did not involve the use of the chemical.