Manuscript received December 27, 2023; revised February 17, 2024; accepted March 21, 2024; published July 5, 2024

Abstract—Stomatal density influences plant photosynthesis, transpiration, and secondary production like fruit and oil. It could serve as a selection criterion for developing plant varieties. The genetic diversity of patchouli is still relatively limited owing to a lack of flowering and fruiting; therefore, genetic variability is also limited. One approach to overcoming this problem is to collect plants from specific regions, called accessions, to identify potential varieties capable of producing abundant and high-quality patchouli oil. Parameters such as stomatal density were evaluated during this process. Conventional manual calculations have inherent drawbacks, including time constraints, low precision, and susceptibility to bias. Therefore, automated methods are essential for stomatal detection models and counting calculations based on deep learning. The dataset consisted of 100 and 400 microscopy images split at a ratio of 8:2 for the training and testing data, respectively. A stomata detection model using YOLOv5 achieved precision, recall, and F1−Score of 0.88 each. The accuracy of the stomata calculation on the test data was 97%. This result demonstrates the ability of the model to calculate the stomatal density in microscopy images.

Keywords—deep learning, patchouli, stomata detection, stomatal density, YOLOv5

Cite: Arie Qur’ania, Yeni Herdiyeni, Wisnu Ananta Kusuma, Aji Hamim Wigena, and Sri Suhesti, "Detection and Calculation of Stomatal Density Using YOLOv5: A Study of High-Yielding Patchouli Varieties," Journal of Image and Graphics, Vol. 12, No. 3, pp. 228-238, 2024.

Copyright © 2024 by the authors. This is an open access article distributed under the Creative Commons Attribution License (CC BY-NC-ND 4.0), which permits use, distribution and reproduction in any medium, provided that the article is properly cited, the use is non-commercial and no modifications or adaptations are made.