Spectral inversion model of the crushing rate of soybean under mechanized harvesting

Rapid and timely acquisition of the crushing rate can help in assessing performance combine harvesters, which is very important for agricultural production. The spectral reflectance soybean provides an alternative method to classical physical chemical analysis laboratory. Therefore, hyperspectral imaging be used rapidly obtain soybean. In this study, was employed, application inter-correlation explored optimization quantitative bands. 130 samples collected from a harvester investigated through Subsequently, raw measured using spectroradiometer equipped with high intensity contact probe under darkroom conditions. Next, (REF) logarithmic reciprocal pretreatment spectrum data (LR) were analyzed compared. effective wavelengths selected according results analysis. Regression models different indices established least squares support vector machine (LS-SVM). inversion model validated compared each other. experimental show that sensitive bands REF are 1061, 1068, 1074, 1090, 2085, 2092, 2095, 2103 nm. Sensitive LR 677, 1039, 1078, 1093, 1101, 1956, 2088, 2107 showed optimal index LS-SVM regression (Rc2 0.939, Rp2 0.915). could not only efficient selection bands, but also retain original sample information. based on realize rapid on-line monitoring (crushing rate) grain harvesters future.