Remote sensing
SatMonitor — Africa Environmental Monitoring
Satellite change detection for deforestation and floods across Africa, from Sentinel imagery to an operator dashboard.
Overview
Built during my Remote Sensing internship at Modelis-France. The platform ingests Sentinel-2 optical and Sentinel-1 radar imagery, computes spectral and SAR change indices between two periods, and flags confirmed versus potential change with a confidence score, image-quality checks and landscape composition. Results are published as map layers and explored in a Vue 3 dashboard where analysts draw an area of interest and launch an analysis.
- Deforestation run over 1,190 km² in Kabylie: 99.2% confidence, 5/5 indices concordant
- Flood run in Lokoja, Nigeria (Sentinel-1): 8.9k ha detected, confirmed vs potential
- Threshold-based change detection owned from design to deployment
Next project
Smart AI Interviewer