Skip to main content
ARS Home » Southeast Area » Dawson, Georgia » National Peanut Research Laboratory » Research » Publications at this Location » Publication #246869

Title: Nondestructive estimation of oil and moisture content using NIR spectroscopy in Valencia and Virginia peanuts

item Sundaram, Jaya
item Kandala, Chari
item Butts, Christopher - Chris
item Holser, Ronald
item Windham, William
item Kays, Sandra

Submitted to: Journal of the American Oil Chemists' Society
Publication Type: Proceedings
Publication Acceptance Date: 9/15/2009
Publication Date: 11/10/2000
Citation: Sundaram, J., Kandala, C., Butts, C.L., Holser, R.A., Windham, W.R., Kays, S.E. 2000. Nondestructive estimation of oil and moisture content using NIR spectroscopy in Valencia and Virginia peanuts. Journal of the American Oil Chemists' Society.

Interpretive Summary: none required.

Technical Abstract: Oil and moisture content of peanuts are important factors in peanut grading. A method by which these parameters could be measured rapidly and nondestructively for peanut pods (in-shell peanuts) would be useful for the industry. In this work, an attempt was made to measure oil and moisture content of Valencia and Virginia type peanuts, without shelling them, using NIR reflectance spectroscopy. Light reflected from peanut samples (about 150 g of pods in each sample, called the calibration sample) at different moisture levels between 6% and 26% was collected in the wavelength range from 400 to 2500 nm using a Foss NIR instrument. Average moisture and oil contents of these samples were determined by standard oven and Soxtec methods, respectively, Fatty acids were measured using gas chromatography to develop standard values. Partial Least Squares analysis was done with different pretreatments on NIR data and respective standard values to develop several prediction models. These models were tested to predict average oil and moisture contents and fatty acids of a different set of samples (validation samples) of peanut pods in the same moisture range of 6% and 26% that were not used in the calibration. The best pretreatment and corresponding model was selected based on the coefficient of determination (R2) and Standard Error of Prediction (SEP). This method was useful in predicting percent oil, moisture, and fatty acids of in-shell Valencia and Virginia market type peanuts.