Skip to content

setup.md

Latest commit

 

History

History
61 lines (42 loc) · 1.08 KB

setup.md

File metadata and controls

61 lines (42 loc) · 1.08 KB

Setup

Needs

  • Python 3.10+
  • Windows, Linux, or WSL2 (WSL2 for ORB-SLAM3)
  • About 8 GB RAM for short clips

Install

cd scope-coverage-map
python -m venv .venv

Windows:

.venv\Scripts\activate
pip install -e ".[dev]"

Linux / WSL:

source .venv/bin/activate
pip install -e ".[dev]"

Test install

pytest tests -v

Run on C3VD

  1. Put data in data/c3vd/trans_t1_a/ (see datasets.md).
  2. python scripts/run_coverage.py --config configs/c3vd.yaml
  3. Open outputs/c3vd/trans_t1_a/coverage.ply in MeshLab.

ORB-SLAM3 (optional)

Default configs use fallback VO (no C++ build). For better SLAM see orb_slam3_wsl.md.

YOLO weights (optional)

mkdir models
python -c "from ultralytics import YOLO; YOLO('yolov8n-seg.pt')"

Or set masking.enabled: false in the YAML.

Extra views

python scripts/export_viewer.py --config configs/c3vd.yaml --c3vd-depth
python scripts/export_gt_coverage.py --config configs/c3vd.yaml
python scripts/verify_run.py outputs/c3vd/trans_t1_a