Kassem Dia, PhD

Postdoctoral researcher in Remote Sensing and Deep Learning, working on the CANOP project in collaboration with INRAE. The project aims at mapping leaf biochemical traits in fruit tree canopies from UAV hyperspectral and LiDAR data, combining 3D physical modeling with machine learning. We focuses on developing a hybrid inversion framework coupling radiative transfer simulations (DART model) with deep learning, applied to high-resolution hyperspectral drone imagery of peach and apricot orchards. Key responsibilities include: • 3D canopy reconstruction from UAV and terrestrial LiDAR point clouds; • Sensitivity analysis and synthetic data generation using the DART radiative transfer model; • Development of deep learning models for the inversion of leaf biochemical traits (chlorophyll, water content) from hyperspectral imagery; • Validation against field measurements (SPAD, DUALEX) acquired during UAV acquisition campaigns.

Phd in image processing and numerical analysis