摘要
探索建立一种有效的大米蛋白质含量近红外光谱检测模型,并寻找1100~2500nm波段中预测大米蛋白质含量的有效波长。采用面积归一化（Area Normalization）方法进行光谱预处理。用主成分回归方法建立回归模型,用Martens不确定性检验方法选择有效波长。发现利用主成分分析可以较好地区分出不同种类的米粉,样品在主成分上的得分可以作为鉴别米粉种类及品质的依据。基于全部波长建立的回归模型,训练集r=0.9923,RMSE=0.0747。交叉验证的结果r=0.9399,RMSE=0.2103。预测集r=0.9364,RMSE=0.1607。基于有效波长建立的回归模型,训练集r=0.9899,RMSE=0.0854。交叉验证结果r=0.9437,RMSE=0.2004。预测集r=0.9079,RMSE=0.1796。使用近红外光谱分析技术检测大米蛋白质含量是可行的,采用Martens不确定性检验方法选择有效波长,并利用有效波长预测大米蛋白质含量也是可行的。
An effective NIR spectral detection model of rice content was explored, and effective wavelengths for the detection of rice protein content between 1100-2500 nm were found. Area normalization method was used to preprocess the spectrum. Principal ~omponents regression （PCR） was used to build the regression model, effective wavelengths were selected by Martens＇ uncertainty test. Different kinds of rice flour could be distinguished well using principal components analysis. Varieties and quality of rice could be identified according to the sample scores along principal components. Regression model based on all wavelengths had the regression coefficient r=0.9923, RMSE=O.0747 in training set, r=0.9399, RMSE=0.2103 in cross validation set, and r=0.9364, RMSE=O.1607 in prediction set. Regression model based on effective wavelengths had the regression coefficient r=-0.9899, RMSE=O.0854 in training set, r=-0.9437, RMSE=0.2004 in cross validation set, and r=-0.9079, RMSE=O.1796 in prediction set. It was feasible to detect rice protein content using NIR spectroscopy. It was also feasible to select effective wavelengths using Martens＇ uncertainty test and to detect rice protein content using these effective wavelengths.
出处
《中国农学通报》
CSCD
2013年第12期212-216,共5页
Chinese Agricultural Science Bulletin
基金
国家自然科学基金"生物聚合物混合光谱检测机理与方法研究"(10664001)
"基于温度和施氮量的水稻品质遥感监测模型研究"(41061039)
"烤烟理化参数的光谱监测机理与方法研究"(11164004)
关键词
大米
蛋白质含量
近红外光谱
主成分回归
有效波长
rice
protein content
NIR spectroscopy
principal components regression
effective wavelengths
