← projectsDTM — Drone LiDAR to Drainage Network Pipeline
End-to-end geospatial pipeline turning raw drone LiDAR into a designed drainage network. Built for the MoPR Geospatial Hackathon at IIT Tirupati.
status completed
stack Python, pysheds, XGBoost, networkx, Geospatial processing
A full geospatial processing pipeline for the MoPR Geospatial Hackathon at IIT Tirupati: ground classification, DTM interpolation, hydrology modeling and drainage network design, running on real drone LiDAR data.
problem
Raw drone LiDAR point clouds are not directly usable for drainage planning — the ground surface has to be extracted, interpolated and modeled hydrologically before any network can be designed.
solution
A staged pipeline: ground classification, DTM interpolation, hydrological flow modeling with pysheds, and drainage network design over the resulting graph.
architecture
- Ground classification from raw drone LiDAR point clouds
- DTM interpolation to a continuous terrain surface
- Hydrological flow modeling with pysheds
- Waterlogging-risk classifier with XGBoost
- Drainage network design with networkx
results
A working end-to-end pipeline from raw drone LiDAR to a designed drainage network, with a waterlogging-risk classifier over real terrain data.