End-to-end pipeline
COLMAP for sparse and dense reconstruction, Open3D for Poisson mesh and vertex colors, then PDAL and GDAL for ground classification, DSM/DTM, and orthophotos. Outputs follow 01_sparse/ … 05_exports/.
GPS-aware workflow
EXIF GPS metadata is read automatically for spatial matching and georeferenced products—many nadir drone projects complete without manual GCPs.
GUI and CLI
Manage projects in the PySide6 desktop UI (map, 3D viewer, GCP editor) or automate with mapfree run and low/medium/high quality presets.
GPU acceleration
NVIDIA CUDA accelerates COLMAP dense stages. Smart Capacity Gate estimates VRAM before dense; CPU fallback remains available.
Organized outputs
Structured directories, per-stage logs under logs/, and YAML presets in mapfree/config/ for reproducible processing.
Quality presets
Low, Medium, and High trade speed vs detail. Hardware VRAM tiers cap image resolution without disabling survey-grade options.
Download builds
Get the latest beta from GitHub releases, with checksums and notes.