基于吸收光谱无损检测水稻叶绿素的研究  

Nondestructive Measurement of the Chlorophyll in Rice by Absorption Spectrum

作　　者：刘子恒[1] 敏乾 谢立强[1] 唐延林[2] 

机构地区：[1]甘肃民族师范学院物理与水电工程系,甘肃合作747000 [2]贵州大学理学院,贵州贵阳550025

出　　处：《甘肃高师学报》2017年第3期35-38,共4页Journal of Gansu Normal Colleges

基　　金：2015年度甘肃省高等学校科研项目"甘肃南部地区野生苦菜叶绿素含量与色彩参数的相关研究"(2015B-123)

摘　　要：采用相关性研究对光谱数据进行了特征光谱筛选,对特征光谱分段处理表征光谱信息,运用主成分分析和多元线性回归的方法构建叶绿素含量的预测模型。实验以水稻活体叶片为研究对象,筛选与叶绿素含量敏感的光谱数据,主要集中在450~550nm和600~700nm两个波段范围,将处理后的特征光谱数据利用多元线性回归和分析的主成分法构建模型.发现485~494nm,505~514nm,515~524nm,650~659nm,670~679nm,690~699nm这6个波段光谱吸收率平均值构建的模型具有较好的效果,定模系数0.921,判定系数0.849,预测相关系数0.883.结果表明此研究方法在处理数据较多的光谱数据构建预测模型方面,不仅简化了模型参数,而且提高了模型的可解释性.The correlation between the chlorophyll content and spectral absorption office leaf were studied. The characteristic spectra which can describe original spectral information have been screened,the spectral model was optimized and explained by the methods of principal component analysis and multiple linear regression.Took rice leaves as the research object.Results show that the overall trend was positively correlated in the vicinity of 450-550nm and 600-700nm,and the chlorophyll content was estimated by the characteristic parameters.The resuhs,a better model was established by six wave bands,there wave bands are 485-494nm,505-514nm, 515-524nm, 650-659nm,670-679nm and 690-699nm.The complex correlation coefficient （R） and the coefficient of determination（R2） are 0.921 and 0.849.and that the correlation coefficient between the estimated values and the real values is 0.883. The results show that the research methods used in this study was simple and efficient.

关 键 词：特征光谱 主成分 预测模型 回归 

分 类 号：S511.5[农业科学—作物学]