This work was conducted to assess the use of Near-infrared reflectance spectroscopy (NIRS) as a technique to analyze nutritional constituents of Distillers dried grain with solubles (DDGS) and corn quickly and accurately, and to apply an NIRS-based indium gallium arsenide array detector, rather than a NIRS-based scanning system, to collect spectra and induce and analyze calibration equations using equipment which is better suited to field application. As a technique to induce calibration equations, Partial Least Squares (PLS) was used, and for better accuracy, various mathematical transformations were applied. A multivariate outlier detection method was applied to induce calibration equations, and, as a result, the way of structuring a calibration set significantly affected prediction accuracy. The prediction of nutritional constituents of distillers dried grains with solubles resulted in the following: moisture (R2=0.80), crude protein (R2=0.71), crude fat (R2=0.80), crude fiber (R2=0.32), and crude ash (R2=0.72). All constituents except crude fiber showed good results. The prediction of nutritional constituents of corn resulted in the following: moisture (R2=0.79), crude protein (R2=0.61), crude fat (R2=0.79), crude fiber (R2=0.63), and crude ash (R2=0.75). Therefore, all constituents except for crude fat and crude fiber were predicted for their chemical composition of DDGS and corn through Near-infrared reflectance spectroscopy.

• Sung Won Choi(Korea Spectral Products) Corresponding author
• Chang Hee Park(Korea Spectral Products)
• Chang Sug Lee(Korea Spectral Products)
• Dong Hee Kim(Korea Spectral Products)
• Sung Kwon Park(National Institute of Animal Science, RDA)
• Beob Gyun Kim(Department of Animal Science and Technology, Konkuk University)
• Sang Ho Moon(School of Food Bio Science, Konkuk University)