An Integrated Method for Tracking and Monitoring Stomata Dynamics from Microscope Videos

Zhuangzhuang Sun, Yunlin Song, Qing Li, Jian Cai, Xiao Wang, Qin Zhou, Mei Huang, Dong JiangPlant 2021 · DOI

TL;DR

This paper introduces the Stomatal Behavior Observation System (SBOS), a novel integrated platform for real-time, non-invasive tracking and analysis of stomatal dynamics in wheat leaves. By combining object tracking and semantic segmentation, SBOS accurately quantifies stomatal opening area (SOA) over time, offering a significant advancement over static image-based methods for understanding plant responses to environmental stress.

What Problem Does This Paper Address?

Plants regulate gas exchange and water transpiration through stomata, pores on their leaf surfaces. Understanding how these stomata open and close, especially in response to environmental changes like drought or chilling, is crucial for plant physiology and agriculture. Historically, researchers relied on static images or indirect measurements, which failed to capture the dynamic, real-time behavior of stomata. A key challenge is the phenomenon of "patchy stomata," where adjacent stomata can exhibit opposite behaviors, leading to complex and heterogeneous responses within a single leaf. Existing methods struggle to capture these dynamic spatial and temporal variations, hindering a deeper understanding of stomatal regulation and its implications for plant stress responses and overall health.

Methods

The SBOS is built around two core modules:

  • Real-Time Observation Module: This module utilizes a portable microscope connected to a laptop to capture high-resolution videos of stomatal dynamics. It features a user-friendly graphical user interface (GUI) for easy operation and allows for simultaneous control of multiple microscopes. Videos are saved with timestamps, facilitating temporal analysis.
  • Automatic Analysis Module: This module performs the heavy lifting of data extraction. It comprises:
  • Object Tracking: Employing a Kernelized Correlation Filter (KCF) algorithm, this component accurately tracks individual stomata across video frames. KCF was chosen for its efficiency and suitability for tracking relatively static targets that only change in size.
  • Semantic Segmentation: A UNet-based deep learning model is used to precisely segment and quantify the stomatal opening area (SOA) in each tracked stoma. The UNet architecture, enhanced with a MobileNet backbone and channel attention mechanisms (SE blocks), is designed for efficient feature extraction and improved accuracy.
  • Graphical User Interface (GUI): A dedicated GUI is provided for the automatic analysis module, making it accessible to researchers without extensive programming expertise.

Datasets

The paper describes the creation of a dataset for training the semantic segmentation model. This dataset consists of stomatal observation videos captured under various conditions (different growth stages, drought, low-temperature, high-temperature). Images were extracted from these videos, cropped to different scales, and annotated. A total of 420 images were annotated, and data augmentation techniques (rotation, flipping, scaling) were applied to expand the training set, resulting in an additional 755 images. The annotated dataset is made publicly available on GitHub.

Key Insights and Ideas

  • Bridging the Gap: SBOS effectively bridges the gap between static observations and dynamic physiological processes, offering a non-invasive way to study stomatal behavior in real-time.
  • Quantifying Patchiness: The system's ability to track and analyze individual stomata allows for the quantitative study of "patchy stomata," revealing distinct differences in initial aperture size and closure rates among stomata, even under uniform stress.
  • Correlation with Gas Exchange: The study demonstrates a strong correlation between quantified stomatal opening area (SOA) and key gas exchange parameters (Pn, Ci, Gs, Tr), suggesting that SBOS can serve as a valuable proxy for these measurements.
  • User-Friendly Tool: The integrated GUI significantly lowers the barrier to entry for researchers, democratizing the study of stomatal dynamics.

Key Results

  • Semantic Segmentation Performance: The best-performing semantic segmentation model (using both data augmentation and SE blocks) achieved a Mean Pixel Accuracy (MPA) of 0.8305 and a Mean Intersection over Union (MIoU) of 0.5590 on the test set.
  • Chilling Stress Response: Under chilling stress, all observed stomata reduced their aperture size and closed within 10 minutes, with notable differences in closure speed between individual stomata (e.g., Stoma 1, 2, and 6 closed faster than Stoma 3, 4, and 5).
  • Drought Stress Correlation: Strong positive correlations were found between mean SOA and gas exchange parameters: Pn (0.93), Ci (0.96), Gs (0.96), and Tr (0.97). This indicates that SOA is a reliable indicator of gas exchange status.

Limitations and Future Work

  • Segmentation Robustness: The authors acknowledge that the robustness of the semantic segmentation module is a key limitation, being dependent on the diversity of the training dataset. The system may struggle with unusual conditions, such as powdery mildew, where bacteria might be mistaken for stomata.
  • Specific Plant Types: Currently, the system is primarily validated for wheat and barley.
  • Future Directions: The authors suggest that further research into the genetic basis of patchy stomata and their role in crop improvement could be facilitated by this system. They also highlight the potential for SBOS to complement traditional gas exchange measurements, offering a more nuanced understanding of plant responses. The open-source nature of the code allows researchers to train their own models for specific plant species or requirements.

Summary updated

Abstract

Patchy stomata are a common and characteristic phenomenon in plants. Understanding and studying the regulation mechanism of patchy stomata are of great significance to further supplement and improve the stomatal theory. Currently, the common methods for stomatal behavior observation are based on static images, which makes it difficult to reflect dynamic changes of stomata. With the rapid development of portable microscopes and computer vision algorithms, it brings new chances for stomatal movement observation. In this study, a stomatal behavior observation system (SBOS) was proposed for real-time observation and automatic analysis of each single stoma in wheat leaf using object tracking and semantic segmentation methods. The SBOS includes two modules: the real-time observation module and the automatic analysis module. The real-time observation module can shoot videos of stomatal dynamic changes. In the automatic analysis module, object tracking locates every single stoma accurately to obtain stomatal pictures arranged in time-series; semantic segmentation can precisely quantify the stomatal opening area (SOA), with a mean pixel accuracy (MPA) of 0.8305 and a mean intersection over union (MIoU) of 0.5590 in the testing set. Moreover, we designed a graphical user interface (GUI) so that researchers could use this automatic analysis module smoothly. To verify the performance of the SBOS, the dynamic changes of stomata were observed and analyzed under chilling. Finally, we analyzed the correlation between gas exchange and SOA under drought stress, and the correlation coefficients between mean SOA and net photosynthetic rate (Pn), intercellular CO<sub>2</sub> concentration (Ci), stomatal conductance (Gs), and transpiration rate (Tr) are 0.93, 0.96, 0.96, and 0.97.
BibTeX
@inproceedings{sun2021integrated,
  author = {Zhuangzhuang Sun and Yunlin Song and Qing Li and Jian Cai and Xiao Wang and Qin Zhou and Mei Huang and Dong Jiang},
  title = {{An Integrated Method for Tracking and Monitoring Stomata Dynamics from Microscope Videos}},
  booktitle = {Plant},
  year = {2021},
  doi = {10.34133/2021/9835961},
}

Loading related content...

Figure