Precision Analytics and Agri-Intelligence (PEARL) Lab

High-Throughput Plant Phenotyping (HTP) & Phenomics
This research focuses on automating crop phenotyping using UAV-based RGB and hyperspectral imaging combined with image analysis and machine learning. It enables high-throughput, non-invasive monitoring of traits such as biomass, canopy structure, and maturity. These tools support precision breeding and large-scale field phenotyping.

Machine Learning & Deep Learning for Agricultural Applications
Advanced machine learning and deep learning models are developed to interpret agricultural imagery and spectral data for trait prediction, segmentation, and classification. Techniques like CNNs and FCNNs have been applied to tasks such as leaf morphology detection and yield estimation, providing robust, scalable solutions for precision agriculture.

Time Series Analysis & Forecasting
This work leverages multi-temporal UAV imagery and sensor data to model and forecast crop development. By combining time series features with ML and deep learning models, it enables accurate prediction of key phenotypic traits such as biomass accumulation and maturity timing across different growth stages.

Remote Sensing & Hyperspectral Data Analysis
Using hyperspectral, multispectral, and thermal imaging, this research extracts detailed biochemical and physiological crop traits. Deep learning and feature selection techniques are employed to detect plant stress, estimate yield, and model spectral signatures—enabling data-driven insights for sustainable crop management.