Growth Signatures of Rosette Plants from Time-Lapse Video
IEEE/ACM Transactions on Computational Biology and Bioinformatics
|December 20, 2015
Summary
This study introduces a novel graph-based tracking algorithm for analyzing early plant development in tobacco plants. The method accurately tracks leaf growth from time-lapse videos, revealing consistent growth patterns in seedlings.
Area of Science:
- Plant biology
- Computational biology
- Image analysis
Background:
- Understanding early plant development is crucial for plant research.
- Analyzing dynamic plant growth from visual data presents challenges due to complex shapes and changing conditions.
Purpose of the Study:
- To develop and validate an automated method for tracking leaf growth in rosette plants using time-lapse imaging.
- To analyze early developmental patterns in Nicotiana tabacum (tobacco) seedlings.
Main Methods:
- Utilized time-lapse videos of Nicotiana tabacum (tobacco) seedlings.
- Employed a novel graph-based tracking algorithm to detect and link potential leaves across video frames.
- Used a leaf-shape model and overlap similarity for leaf detection and tracking, handling occlusions and appearance changes.
Main Results:
- Successfully tracked the first leaves of tobacco plants, including partially or temporarily occluded ones.
- Demonstrated the method's applicability for automatic plant growth analysis.
- Identified consistent growth behavior and a characteristic growth signature across all analyzed seedlings.
Conclusions:
- The developed graph-based tracking algorithm effectively analyzes early plant growth from time-lapse videos.
- The method provides a robust solution for automatic plant growth analysis, even with challenging imaging conditions.
- Consistent growth signatures were observed in early tobacco seedling development.
