PREDICTION OF CORN DRY-MILLING QUALITY BY NEAR-INFRARED SPECTROSCOPY

The potential for near-infrared (NIR) reflectance spectroscopy to pre- diet corn dry-milling quality has been investigated. NIR spectra were obtained from samples of whole-kernel yellow dent corn. Dry-milling qu ality of the samples was evaluated with a Tangential Abrasive Dehullin g Device (TADD), and by a short-flow laboratory milling procedure that allowed calculation of a milling evaluation factor (MEF). Calibration s relating TADD index and MEF to NIR measurements were developed and t ested with validation sample sets. The results indicate that NIR spect roscopy can predict dry-milling characteristics with a reliability sui table for at least rough screening. For MEF, the best calibration yiel ded a correlation coefficient of 0.90 and a standard error of predicti on (SEP) of 1.55%. SEP values for TADD index were somewhat higher. Der ivatization of the NIR spectra lowered the SEP values compared to the use of underivatized log 1/R measurements.