0.818,、0.069%和0.085%,。Spectral properties of simply treated soil samples were analyzed by using Nicolet intelligent Fourier transform (FT) infrared spectrum. A new pretreatment method-orthogonal signal correction (OSC) was presented to eliminate the influence of the noise on soil organic matter (SOM) content prediction. Partial least square (PLS) analysis has been used to build prediction models with calibration data of 67 samples. The remaining 20 samples were used to validate the models. The result showed that OSC-PLS could improve the prediction ability greatly. The correlation coefficient is 0.893, standard error of prediction (SEP) is 0.051%, and root mean standard error of prediction (RMSEP) is 0.050% respectively.
參考文獻(xiàn)
相似文獻(xiàn)
引證文獻(xiàn)
引用本文
宋海燕,何勇.基于OSC和PLS的土壤有機(jī)質(zhì)近紅外光譜測(cè)定[J].農(nóng)業(yè)機(jī)械學(xué)報(bào),2007,38(12):113-115.[J]. Transactions of the Chinese Society for Agricultural Machinery,2007,38(12):113-115.