毛豆叶面积回归测算及统计分布研究

Study on Regression Estimation and Probability Distribution of Leaf Area for Soybean

在农业灾害研究、农田灌溉估算、作物栽培指导等实际应用中都需要测算叶面积。通过MATLAB程序对数码相机拍摄的毛豆叶片照片进行处理,计算得出叶长、叶宽和叶面积,并分别对叶面积与长乘宽、叶面积与长的平方的关系进行拟合,发现叶面积与长乘宽的拟合方程精度较高。研究了叶片指标的统计分布规律,并采用K-S(Kolmogorov-Smirnov)检验法对可能的分布进行了检验。结果发现对数正态分布和Gamma分布较为符合叶片指标的实际分布。给出了毛豆的叶面积的实际测量方法,并为叶面积的研究和应用提供了理论依据。

Estimation of the leaf area is essential for agriculture disaster research,irrigation estimation and crop cultivation.In this paper,the leaf area,leaf length and leaf width of soybean are processed by MATLAB and the relationship among the leaf area,leaf length and length width is studied.Several regression functions are found and verified to be useful in the calculation of leaf area.The probability distribution of leaf index is studied and the lognormal model and Gamma model are demonstrated to fit the real distribution of leaf area.