<![CDATA[Leaves are essential for photosynthesis and plant productivity, making leaf measurement critical for assessing growth and health. Traditional measurement techniques provide useful and interpretable results but are time?consuming and risk damaging fragile leaves. This study aims to combine digital image analysis with numerical integration method, particularly Newton-Cotes algorithms including the Riemann sums, the composite trapezoid rule, Simpson’s 1/3 and 3/8 rules, Boole’s rule, and the six point NC rule for leaf area estimation. The leaf area is estimated by partitioning and summing the area of curve?bounded objects from different number of subdivisions. A total of 86 leaf samples were collected, and their accuracy and agreement were compared against the grid method. Pearson correlation coefficient was employed to assess the strength of the relationship, while paired samples t-test and Bland–Altman plot was used to evaluate the mean difference and the level of agreement between the two methodologies, respectively. Our findings revealed that higher partition sizes enhanced accuracy, reduced integration error, lowered RMSE, and produced estimates highly consistent with reference leaf area, supported by a Pearson correlation coefficient (r) of 0.999 and non?significant t-test results (?>0.05). The Bland–Altman analysis confirmed narrower variability at higher partitions resulting near?perfect agreement, particularly under the Riemann rule. Finally, leaf area estimation using digital image analysis combined with Newton–Cotes algorithms provide accurate, efficient, and reliable results, demonstrating strong agreement, statistical consistency, and practical scalability for leaf analysis.]]>
Copyrights © 2026
