<![CDATA[This study aimed to develop a platform to estimate leaf sizes based on digital vision-based image processing. The developed approach was capable of quantifying the sizes of leaves more easily and more accurately compared to traditional measurement. Python, coupled with a PyTQ5 module, was utilized to develop the platform. Acquired leaf images were processed to extract size features, consisting of area, perimeter, length, and width. The actual sizes were measured by placing the leaf samples on a millimeter graph paper and counting the grids inside the leaf’s edge. The estimated leaf sizes were predicted using image analysis by calibrating pixels per mm in the image. Simple linear regression was performed to check the strong relationship between the actual and predicted sizes. For all size parameters, the obtained R2 reached very close to 1.0, while the obtained RMSEs were lower than 0.10. According to Bland-Altman (B&A) plots, the differences between the measured and predicted sizes are normally distributed, and almost all the points lie inside the 95% limit of agreement. Finally, our proposed platform could be an alternative to estimate leaf sizes, which provides rapid, accurate, and non-destructive measurement]]>
Copyrights © 2026
