基于图像处理的玉米叶色分级及其与光合性状的相关性分析  

Classification of Maize Leaf Color Based on Image Process and the Relationship with Photosynthetic Traits

作　　者：李长生[1] 王国宏[2] 姜敏[2] 孙权[1] 张春宇[1] 林凤[1] 

机构地区：[1]沈阳农业大学生物科学技术学院,沈阳110161 [2]辽宁省农业科学院玉米研究所,沈阳100161

出　　处：《沈阳农业大学学报》2012年第4期411-417,共7页Journal of Shenyang Agricultural University

基　　金：国家自然科学基金项目(35071266)

摘　　要：以67份辽宁省玉米骨干自交系为材料,利用图像处理技术提取叶片颜色特征值,构建叶色分级系统,并对颜色特征值与玉米叶片叶绿素含量及光合速率进行相关性分析,建立回归模型。结果表明:综合全部颜色特征值,可将67份玉米自交系叶色划分为6个等级;GB组合可以使其与叶绿素含量间的相关性提高;G/B、G-B、2G-R-B、(G-B)/(G+B)、b、b*与叶绿素含量呈显著相关,G-B模型预测的误差值最小,为13.48%。颜色特征值G-B可作为基于图像处理测定玉米叶绿素含量的最佳预测指标,其所预测的叶绿素a、b、a+b含量模型分别为:y=-0.201x+9.549,y=-0.062x+3.143,y=-0.263x+12.692。Leaf color, as a main phenotypic trait, is widely used in identification of maize gerrnplasm resources and trait investigation of variety tests. In addition, leaf color is related with chlorophyll contents which affect photosynthetic rate. In the present study, leaf color characteristics were dealt with image process technology, a leaf color recognition system was established based on 67 backbone maize inbred lines in Liaoning Province. The correlations among color characteristic, chlorophyll contents and photosynthetic rate were analyzed using the regression models. The results suggested that 67 maize inbred lines were classified into 6 groups considering all the color characteristic value. Combination of GB enhanced its correlation with chlorophyll content; G/B, G-B, 2G-R-B, （G-B） / （G＋B）, b, b^＊ showed significant correlations with chlorophyll contents; model G-B showed the lowest prediction error （13.48%）. The color of G-B was the best index to predict maize chlorophyll contents based on image process. The formula of model G-B for predicting chlorophyll contents was y=-0.201x＋9.549, y=-0.062x＋3.143 and y=- 0.263x＋12.692 for chlorophyll a, b and a＋b, respectively.

关 键 词：玉米 图像处理 颜色特征值 叶绿素含量 光合速率 

分 类 号：S326[农业科学—作物遗传育种]