Abstract:After being treated by gamma-ray, the spectral characteristic of rapeseed would be changed. Based on the principle, a rapid and nondestructive method by using visible and near infrared spectroscopy was proposed to discriminate rapeseeds (Brassica nupus) treated by different dosages of gamma-ray. Partial least square (PLS) method and BP neural network (BPNN) were applied to establish the discrimination model, and the influences of different pretreatment methods of original spectra data, data transformation methods of PLS principal components and the selection of node number of hidden layers of BP neural network model on prediction precision were compared and discussed. In the experiment, 184 samples were treated by gamma-ray with 5 different dosages (50, 100, 150, 200 Gy, and the samples without gamma-ray treatment). Then spectra tests were performed on the 184 samples using a spectrophotometer (325-1 075 nm). One hundred thiry five samples were selected randomly for model calibration and the left 49 samples were used for prediction. As a result, the optimal model was established and the parameters of the model were shown as follows. The original spectra data were pretreated by smoothing media filter, multiplicative scatter correction and Savitzky-Golay derivatives, then 6 PLS principal components were selected by using partial least square method. After being transformed by using natural logarithm transformation method, the 6 PLS principal components were used as the input layer factors to establish the BP neural network model and the node number of hidden layers was selected as 4 or 9. The prediction precision of the optimal model to distinguish the untreated samples from gamma-ray treated samples was 100%. The precision of predicting the dosages of gamma-ray treatment of all samples achieved 85.71%. It can be concluded that the proposed method for estimating the influence of different gamma-ray dosages on the spectral characteristic of treated rapeseeds was feasible.
Key words:Vis/NIR spectroscopy;Brassica napus;Gamma-ray;Partial least square;BP artificial neural network
黄敏1,王遵义2,何勇1*. 辐照处理对油菜籽光谱特性影响的研究[J]. 光谱学与光谱分析, 2008, 28(11): 2540-2544.
HUANG Min1,WANG Zun-yi2,HE Yong1*. Study on Influence of Gamma-Ray Treatment on Spectral Characteristic of Rapeseed. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2008, 28(11): 2540-2544.