I am an MSc graduate in Geoinformation Science and Earth Observation at the University of Twente (ITC) and Lund University, worked at the intersection of UAV imaging, deep-learning segmentation, and applied Earth Observation for agriculture.
- Completed my MSc thesis on UAV-based crop mapping and fractional vegetation estimation in smallholder landscapes of Mozambique, using a Segformer transformer architecture.
- Worked as a Geospatial Intern at Ramani B.V. (The Netherlands), validated Earth Observation implementations of Verra's VM0047 methodology for nature-based carbon removal and contributed to AI-driven agricultural advisory systems.
- UAV and deep-learning phenotyping for crop and disease monitoring
- Earth Observation at scale (Sentinel, Landsat, MODIS)
- Agricultural AI and decision-support systems for growers and breeders
- Spatial statistics and validation of remote-sensing products
- Climate-smart agriculture
Languages & ML: Python (PyTorch, scikit-learn, geopandas, xarray), MATLAB Geospatial: Google Earth Engine, ArcGIS, QGIS, PIX4D, TIMESAT Deep learning: Segformer and related transformer architectures, semantic segmentation Statistics: Theil–Sen, Mann–Kendall, trend analysis, accuracy assessment Process-based simulation modelling: LPJ-GUESS dynamic vegetation model
- 🌐 Portfolio: https://collins-67.github.io/
- 💼 LinkedIn: collins-edem-hlordzie
- 📧 Email: edemcollins67@gmail.com
Note: some of my current research code lives in institutional and privately governed repositories. Some of the public repositories here are illustrative of methods and workflows I engage with in that broader work.