Estimating leaf area of Camelina through non-destructive models

This study focuses on quantifying Camelina leaf area based on leaf dimensions and selecting the best models. The experiment assessed Camelina leaf area under water stress, with treatments comprising five irrigation levels (۱۰۰%, ۹۰%, ۷۵%, ۶۰%, and ۵۰% of field capacity), each replicated three times. The analysis utilized SAS software for checking data normality and generating scatter plots. Additionally, Excel's Solver and Regression Learner Toolbox of MATLAB were used to fit models to the experimental data. A relationship was developed using length (L), width (W), and the product of length by width (LW) as independent variables and leaf area as the dependent variable. Results indicated that models based on LW presented more accurate results, while regression learner models based on width provided the best overall performance. Eq. (۱۰) outperformed other models. According to MATLAB, Linear Regression provided the best results for LW-LA and L-LA, whereas Gaussian Process Regression presented superior performance for W-LA.

کلیدواژه ها:

leaf area ، leaf length ، regression learner ، water stress