Use of Nir Spectroscopy and Ls-Svm Model for the Discrimination of Varieties of Soil

Near-infrared spectroscopy (NIR) combined with chemomatrics method wasemployed in this paper. The objectives were to investigate the feasibility ofusing VIS/SWNIR spectroscopy to discriminate soil of different kinds, and tovalidate the performance of selected sensitive ebands. Spectrums in the NIRregions (4003.563-12496.67 cm-1) were collected from 380 samples and thedata was expressed as absorbance, the logarithm of the reciprocal ofreflectance(log 1/R).240 samples were randomly collected as modeling, andqualitative analysis for the discrimination, but for the sake of speedup theclassification and 10-folds cross-validation were used to model, and Based onclassification is apparently more effective. At the same time, the sensitivebands of soil varieties were caculated,in which we found the 801-972 nm can99% indicated that the NIR can mainly represent the characteristics of soil ofdifferent kind based on SVM model.