Prediction of sorghum oil content using near‐infrared hyperspectral imaging

Abstract

Background and Objectives Aside from being a staple crop, sorghum is now being used as a gluten-free food, an animal feed ingredient, and a biofuel source. This growing demand for sorghum has increased interest in grain quality and utilization. This study explored near-infrared hyperspectral imaging (NIR HSI) as a non-destructive and rapid method to predict the oil content of sorghum grains. Findings Partial least square (PLS) regression models for oil from NIR HSI spectra achieved 0.19% standard error of calibration (SEC) and 0.21% standard error of prediction (SEP) at 10 PLS factors. The results from the NIR HSI instrument were comparable to those from the single-kernel near-infrared reflectance instrument (SKNIR) using the same set of samples. Conclusion This study showed the potential of hyperspectral imaging as a quality control method for sorghum grains, specifically for oil content, which could be beneficial for sorghum breeders, growers, and processors. Significance and Novelty The increasing interest in sorghum use prompted this study which is one of the first to explore NIR hyperspectral imaging for sorghum oil with the ability to indicate single seed weight. This article is protected by copyright. All rights reserved.