The article considers methods of identifying the varieties of the components of grain mixtures based on the spectral analysis in the visible and near-infrared wavelength ranges. Various measurement approaches - reflection, transmission and combined spectrum methods - are used. The results of the spectral measurement are processed using neural network based classification algorithms combined with data dimensionality reduction techniques. The probabilities of incorrect recognition for various numbers of features and combinations of spectral methods are estimated for mixtures of wheat grains. Combined use of spectral methods allowed to reduce the classification error by about ten times and bring its absolute value to 0.01...0.001.
展开▼
