Related Experiment Video
Updated: Oct 24, 2025
06:41
Imaging and Analysis for Quantifying Maize (Zea mays) Abiotic Stress Phenotypes
Published on: March 28, 2025
1.2K
An Affordable Image-Analysis Platform to Accelerate Stomatal Phenotyping During Microscopic Observation
Yosuke Toda1,2,3, Toshiaki Tameshige4,5, Masakazu Tomiyama2
1Japan Science and Technology Agency, Saitama, Japan.
Frontiers in Plant Science
|August 16, 2021
Summary
We developed a low-cost, real-time stomata detection platform using computer vision for plant phenotyping. This system enables rapid, on-site analysis of stomatal traits like density and size in wheat.
Area of Science:
- Plant Biology
- Computer Vision
- Agricultural Science
Background:
- Image-based quantification of stomatal traits is advancing rapidly.
- High installation costs and operational complexity hinder on-site application for biologists.
Purpose of the Study:
- To present a low-cost, real-time stomata detection platform for microscopic observation.
- To overcome the limitations of existing expensive and complex phenotyping systems.
Main Methods:
- Developed a system integrating a deep neural network stomata detector with an upright microscope, USB camera, and GPU-enabled single-board computer.
- Utilized commercially available hardware and free cloud services for model training.
- Trained the model to detect dumbbell-shaped stomata in wheat leaf imprints.
Main Results:
- The platform enables real-time stomata detection during microscopic observation.
- Collected comprehensive stomatal phenotypes from wheat leaves, confirming differences in stomatal density (SD) and stomatal size (SS) between surfaces and species.
- Demonstrated the system's effectiveness in identifying variations in stomatal traits.
Conclusions:
- The proposed low-cost platform facilitates accessible, real-time stomata phenotyping.
- This technology is expected to accelerate research in plant physiology and breeding by enabling efficient stomatal trait analysis.
