vision-ai-service

Analyserer en videostrøm for passeringer (people crossing line). Tar skjermbilde av passeringen.

Start service: python3 -m vision_ai_service.app But first, start dependencies (services & db): docker-compose up event-service user-service photo-service mongodb

Development Requirements

• pyenv
• pipx
• poetry
• nox
• nox-poetry
• pillow

If required - virtual environment

Install: curl https://pyenv.run | bash Create: python -m venv .vivenv (replace .venv with your preferred name) Activate:source .vivenv/bin/activate

Install

% git clone https://github.com/heming-langrenn/vision-ai-service.git % cd vision-ai-service % pyenv local 3.11 % poetry install

### Prepare .env filer (dummy parameter values supplied)
LOGGING_LEVEL=INFO
JWT_SECRET=secret
JWT_EXP_DELTA_SECONDS=3600
ADMIN_USERNAME=admin
ADMIN_PASSWORD=password
DB_USER=admin
DB_PASSWORD=password
EVENTS_HOST_SERVER=localhost
EVENTS_HOST_PORT=8082
PHOTOS_HOST_SERVER=localhost
PHOTOS_HOST_PORT=8092
USERS_HOST_SERVER=localhost
USERS_HOST_PORT=8086

### Run all sessions

% nox

### Run all tests with coverage reporting

% nox -rs tests

### Push to docker registry manually (CLI)
docker-compose build
docker login ghcr.io -u <github username>
password: Use a generated access token from GitHub
docker tag ghcr.io/langrenn-sprint/vision-ai-service:test ghcr.io/langrenn-sprint/vision-ai-service:latest
docker push ghcr.io/langrenn-sprint/vision-ai-service:latest