2026-09-03 –, Room3
—An Observation-Aware, Offline-First Production System from UAV Acquisition to Decision-Ready Products—
This talk presents MIMAR, an observation-aware production architecture built around OpenDroneMap and OpenSfM for large-scale tropical plantations. It connects controlled UAV acquisition, high-recall SfM in repetitive canopies, GPU-accelerated matching and bundle adjustment, recoverable offline execution, geospatial QA, and tree-level decision products.
The focus is not a single faster algorithm, but an end-to-end system that remains reproducible and operational under field, compute, and business constraints.
MIMAR is an independent, offline-first production architecture built around OpenDroneMap, OpenSfM, OpenMVS, and the wider FOSS4G ecosystem for difficult UAV missions in large tropical plantation settings.
Repetitive canopies, uncertain pose, and changing field conditions create ambiguous matches and fragile multi-view tracks. MIMAR controls solar time, illumination, weather, GSD, overlap, and sidelap as an input contract before processing. Position and pose act as priors; image evidence, robust multi-view geometry, and bundle adjustment determine reconstruction. High-recall GPU matching and accelerated bundle adjustment support that evidence-preserving path.
Raw imagery is processed locally in reproducible Docker environments with QA gates, evidence logs, and recovery states. Only validated geospatial and tree-level products move when connectivity is available.
MIMAR is not a new SfM algorithm. Its contribution is a controlled production architecture that makes decision-ready delivery repeatable under field, compute, and business constraints.
Helpful background—not required—is familiarity with Structure-from-Motion (SfM), Multi-View Stereo (MVS), local feature matching, and geospatial point-cloud and raster workflows. The talk introduces MIMAR-specific controls and acceleration paths in context.
Recommended starting points:
- OpenDroneMap documentation: https://docs.opendronemap.org/
- OpenSfM documentation: https://opensfm.org/docs/
- OpenMVS project and documentation: https://github.com/cdcseacave/openMVS
- FAISS documentation: https://faiss.ai/
- PDAL documentation: https://pdal.io/
- GDAL Cloud Optimized GeoTIFF documentation: https://gdal.org/en/stable/drivers/raster/cog.html
OpenDroneMap (ODM), OpenSfM, OpenMVS, PopSift, FAISS, PDAL, GDAL, and Docker
I make my conference contribution available under the CC BY 4.0 license. The conference contribution comprises the abstract, the text contribution for the conference proceedings, the presentation materials as well as the video recording and live transmission of the presentation:
A geospatial and remote sensing engineer, specializing in scalable and viable systems that integrate product design, production engineering, and operational efficiency.