94.74%、94.44%、92.31%。 The nondestructive discrimination of the shelled chestnuts was studied with near infrared spectra, which could provide a new method for quality detection of other shelled agricultural products. 178 chestnut samples were prepared and the diffuse spectral reflectance of the samples were collected in the wave number range of 12000~4000cm-1. First,six preprocessing methods including smooth、vector normalization、min-max normalization、standard normal variate transformation、multiplication scattering correction and first derivative were used to improve the original spectrum. Then,principal component analysis was applied to compress thousands of spectral data into several variables and to collect spectral information. The principal components extracted by PCA were employed as the inputs of the BP neural networks. Effects of the six preprocessing methods on the models based on BP neural network were compared. The results show that prediction precision varied to different preprocessing methods. The optimum network structure of 7-4-1 was obtained after vector normalization method. Discriminating rate of qualified chestnut, surface moldy chestnut and internal moldy chestnut in prediction reached 94.74%, 94.44% and 92.31%, respectively.
参考文献
相似文献
引证文献
引用本文
周竹,刘洁,李小昱,李培武,王为,展慧.霉变板栗的近红外光谱和神经网络方法判[J].农业机械学报,2009,40(Z1):109-112. of Moldy Chinese Chestnut Based on Artificial Neural Network and Near Infrared Spectra[J]. Transactions of the Chinese Society for Agricultural Machinery,2009,40(Z1):109-112.