基于NIR的模式识别技术在地理标志产品响水大米鉴别中的应用

Identification of geographical indication product Xiangshui-rice based on NIR and pattern recognition technology

分别利用多元线性回归判别分析和BP人工神经网络分析建立了近红外光谱（NIR）快速鉴别地理标志产品响水大米的新方法．大米的近红外光谱数据经过一阶导数和平滑处理后，利用主成分分析对数据进行了降维处理，并确定了相关性最大的特征波段（7700～6700cm-1与5700～4300cm-1）．利用特征波段的主成分数据建立了多元线性回归判别分析和BP人工神经网络鉴别模型．2种模型对于地理标志产品响水大米的鉴别正确率均为100％，适用于地理标志产品的快速无损鉴别．

A rapid method was developed for discrimination of geographical indication Xiangshui-rice with discriminate multiple linear regression （DMLR） and BP artificial neural network（BP-ANN）by near infrared spectroscopy （NIR）. After first derivative and smooth processing, principal component analysis （PCA） was used to reduct the dimensionality of the spectral datas. And characteristic wave band （7 700- 6 700 cm-1 , 5 700-4 300 cm-1） with max-relativity was determined. The discrimination methods of Xian- gshui rice by DMLR and BP-ANN were established based on PCA data in characteristic wave band. The correct rate of identification by the two methods were all 100%, which were suitable to discrimination of geographical indication product through fast and nondestructive method.