基于透过强度光谱红边特征的辣椒叶片叶绿素含量诊断模型研究  

Research on the diagnostic model of chlorophyll content in pepper leaves based on the red-edge characteristics of transmission intensity spectra

作　　者：黄帅 孙浩翔 王廷超 舒坤良[1] 王青蓝[1] 盖嘉慧[1] Huang Shuai;Sun Haoxiang;Wang Tingchao;Shu Kunliang;Wang Qinglan;Gai Jianghui(Jilin Academy of Agricultural Sciences,Changchun,130033,China)

机构地区：[1]吉林省农业科学院,长春市130033

出　　处：《中国农机化学报》2023年第11期109-114,共6页Journal of Chinese Agricultural Mechanization

基　　金：吉林省农业科技创新工程项目(CXGC2023DX006);吉林省农业科技创新工程项目(CXGC2021TD019);吉林省科技发展计划项目(20200402003NC)。

摘　　要：为快速、精准的诊断辣椒叶绿素含量,采用测量方法简单、干扰因素较少的透过光谱进行定量分析。研究对叶绿素含量极其敏感的红边波段变化规律,利用一阶导数法计算代表红边波段特征的红边位置、红边幅值和红边面积三种红边参数,分析参数变量间的相关性和共线性,并基于特征参数分别建立最小二乘回归模型和岭回归模型。结果表明,最小二乘回归模型的验证结果相对误差为7.94%,岭回归模型的验证结果相对误差为7.75%,岭回归模型不仅解决参数间的共线性问题,并且提高模型验证结果的精确度,该模型可以作为辣椒生长状态快速诊断的技术基础。In order to quickly and accurately diagnose the chlorophyll content of peppers,quantitative analysis was carried out through the transmission spectrum with simple measurement method and less interference factors.The variation law of the red-edge band,which was extremely sensitive to chlorophyll content,was studied.Three red-edge parameters(λred),red-edge amplitude(dλred)and red-edge area(Sred)representing the characteristics of the red-edge band were calculated by the first derivative method.The correlation and collinearity among the parametric variables were analyzed.And the least squares regression model and the ridge regression model were established respectively based on the characteristic parameters.The results showed that the relative error of the verification results of the least squares regression model was 7.94%,and the relative error of the verification results of the ridge regression model was 7.75%.The ridge regression model not only solved the problem of collinearity between parameters,but also improved the accuracy of the model verification results.Therefore,the above model can be used as a technical basis for rapid diagnosis of pepper growth status.

关 键 词：辣椒叶片 叶绿素含量 透过强度光谱 红边参数 岭回归 

分 类 号：S641.3[农业科学—蔬菜学]