Estimating Tea Plant Physiological Parameters Using Unmanned Aerial Vehicle Imagery and Machine Learning Algorithms

Area of Science:

• Agricultural Science
• Remote Sensing
• Plant Physiology

Background:

• Climate change threatens tea production, impacting growth, photosynthesis, yield, and quality.
• Accurate real-time monitoring is crucial for effective tea plantation management and production stability.

Purpose of the Study:

• To develop and validate predictive models for estimating tea plant physiological parameters using remote sensing data.
• To compare the effectiveness of conventional farming methods (CFMs) and agroecological farming methods (AFMs) in central Taiwan.

Main Methods:

• Integrated leaf area index (LAI), photochemical reflectance index (PRI), and quantum yield of photosystem II (ΦPSII) data.
• Utilized unmanned aerial vehicle (UAV)-derived visible-light and multispectral imagery to compute color and multispectral indices (MIs).
• Employed feature ranking and eight regression algorithms, including XGBoost, to estimate physiological parameters.

Main Results:

• Agroecological farming methods (AFMs) showed lower LAI but greater environmental adaptability (higher PRI and ΦPSII) compared to conventional farming methods (CFMs).
• Multispectral indices (MIs) provided stable predictions for tea plant physiological parameters.
• XGBoost demonstrated superior performance in predicting LAI (R²=0.716), PRI (R²=0.643), and ΦPSII (R²=0.920).

Conclusions:

• Integrating gradient boosting models with multispectral data effectively captures tea plant physiological characteristics.
• Developed generalizable predictive models advance non-contact crop monitoring for precision tea plantation management.
• Findings provide a scientific foundation for enhancing tea crop resilience and sustainable agricultural practices.