BjaeverMetrics

BjaeverMetrics is an advanced football analysis program designed for IF Frem Bjæverskov. It is used by coaches for advanced statistic to build winning strategies.

Run pip install -r requirement.txt

Local web app

Install dependencies:

pip install -r requirements.txt

Start the local API and frontend:

uvicorn bjaevermetrics_app:app --reload --host 127.0.0.1 --port 8000

Open http://127.0.0.1:8000 in the browser.

The local app serves frontend/, receives video uploads at POST /api/analyze, runs the pipeline on the local machine, stores results in SQLite, and shows saved matches through GET /api/matches.

BJEAVERMETRICS/ ├── .venv/ ├── .gitignore ├── pyproject.toml ├── requirements.txt ├── requirements-dev.txt ├── README.md │ ├── configs/ │ ├── default.yaml # default runtime config │ └── test.yaml # overrides for --test mode │ ├── data/ # gitignored (large files) │ ├── raw/ # original Veo MP4s │ ├── annotated/ # your labeled frames for fine-tuning - training data/benchmarking data │ ├── interim/ # detections.json, tracking.json per match │ ├── processed/ # final stats.json/parquet │ └── models/ # best.pt, pretrained weights │ ├── src/ │ └── fodbold/ # your package │ ├── init.py │ ├── config.py # load + validate config │ ├── cli.py # argparse/typer logic (called by main.py) │ │ │ ├── io/ │ │ ├── init.py │ │ ├── video.py # OpenCV reader, frame iterator │ │ └── metadata.py # holdnavne, halvleg, hjemme/ude │ │ │ ├── detection/ │ │ ├── init.py │ │ └── detector.py # YOLO wrapper │ │ │ ├── tracking/ │ │ ├── init.py │ │ └── tracker.py # ByteTrack/BoT-SORT wrapper │ │ │ ├── stats/ │ │ ├── init.py │ │ ├── possession.py │ │ ├── touches.py │ │ └── passes.py │ │ │ └── pipeline.py # orchestrates the stages │ ├── tests/ │ ├── init.py │ ├── test_video.py │ ├── test_detector.py │ └── fixtures/ # tiny test clips │ ├── notebooks/ # spikes, model comparisons, exploration │
│ ├── scripts/ # one-off utilities │
│ └── main.py # thin CLI wrapper → src/fodbold/cli.py