On-line measurement of grain quality with NIR technology

A feasibility study was conducted to measure grain quality (moisture and protein content) with near-infrared reflection (NIR) technology on a conventional New Holland TX64 combine harvester. A Zeiss Corona NIR sensor was installed on a bypass unit of the clean grain elevator, and reference samples were taken at the end of the bubble-up auger, which transports the grain from the clean grain elevator to the grain bin. Measured signals at the diode array of the spectrophotometer were highly variable in time, and an appropriate low-pass filter was designed to extract relevant spectra at a sampling rate of 1 Hz. In addition, an optimal time shift was calculated between the location of the NIR measurement and the reference sampling to guarantee an optimal synchronization between the measured spectra and reference values. Preprocessing, calibration, and validation were executed using the PLSplus/IQ module of the GRAMS/32 software package. The calibration models were developed using the PLS algorithm and validated through cross-validation. Cross-validation errors of 0.57% and 0.31% were achieved for the on-line protein and moisture content measurements, respectively. These accuracies are comparable with the prediction accuracy that was achieved on a similar, but stationary, unit and prove the technical feasibility of estimating grain quality on-the-go with NIR technology.