This research aimed to propose an online system based on multispectral images for the real-time estimation of the moisture content (MC) of sugarcane bagasse. The system consisted of a conveyor belt, four halogen bulbs, and a multispectral camera. The MC models were developed using machine learning algorithms, i.e., multiple linear regression (MLR), principal component regression (PCR), artificial neural network (ANN), PCA-ANN, Gaussian process regression (GPR), PCA-GPR, random forest regression (RFR), and PCA-GPR. The models were developed using 150 samples (calibration set) meanwhile the remaining 50 samples were applied as a validation set. The comparison of all developed models showed that the PCA-RFR model achieved better detection with a higher accuracy of MC prediction. The PCA-RFR model showed the best results which were a coefficient of determination of prediction (r2) 0.72, root mean square error of prediction (RMSEP) 11.82 wt%, and a ratio of the standard error of prediction to standard deviation (RPD) of 1.85. The results show that this technique was very useful for MC rapid screening of the sugarcane bagasse.

On-line measurement of cane bagasse is needed for its utilization. This study aims to investigate effect of the measurement condition of multispectral image for on-line systems. Two main factors (belt speeds and installation angles of light source) influencing the on-line image collection were studied. Four and two levels of the belt speeds (5, 10, 15, and 20 cm/s) and the installation angles (30 and 45°) were specified in this work, respectively. The effect of these two factors on image repeatability was analyzed by ANOVA using a significant level of 5%. The result showed that belt speeds and angles of light source did not influence image repeatability. However, we found that light source angle of 30 degree gave higher precision than 45 degrees