video-service

A service to capture relevant video clips from an online camera. The service can run as worker in 2 modes:

CAPTURE_LOCAL

Capture video from a stream using Python libraries (cv2.VideoCapture), save as video clips in configurable resolution and duration. Video is stored locally and then uploaded to cloud bucket.

DETECT

Line crossing detection. The service run as stand alone worker and pick up videos from the cloud bucket. Configuration from database (default values in global_settings.json) will always be shared between the workers while each worker's mode will be defined through environment (env) configuration.

storage settings

Sets usage of local file storage or cloud services such as Buckets valid storage modes (VIDEO_STORAGE_MODE): local_storage - pushing video to cloud bucket cloud_storage - pushing image detectsions to cloud bucket

Requirement for development

Install uv, e.g.:

% curl -LsSf https://astral.sh/uv/install.sh | sh

Then install the dependencies:

% uv sync

If required - virtual environment

curl https://pyenv.run | bash
python -m venv .venv
pyenv install 3.13
source .venv/bin/activate

Start service in virtual env:

set -a
source .env
set +a
python -m video_service.app
Dependencies (services & db):
docker compose up integration-service race-service competition-format-service photo-service user-service event-service mongodb photo-service-gui

Install

% git clone https://github.com/heming-langrenn/video-service.git % cd video-service

Prepare .env filer (dummy parameter values supplied)

LOGGING_LEVEL=INFO ADMIN_USERNAME=admin ADMIN_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 GOOGLE_CLOUD_PROJECT=sigma-celerity-257719 GOOGLE_STORAGE_BUCKET=langrenn-sprint GOOGLE_STORAGE_SERVER=https://storage.googleapis.com GOOGLE_CLOUD_REGION=europe-north1 MODE=CAPTURE_LOCAL

Valid MODE values: CAPTURE_LOCAL, DETECT

CAPTURE_LOCAL: Traditional Python video capture from URL

DETECT: Line crossing detection

EVENT_ID: Optional - if several events are defined in photo-service db

Running tests

We use pytest for contract testing.

To run linters, checkers and tests:

% uv run poe release

To run tests with logging, do (/home/heming/Nedlastinger/20250525_GKOpp1.mp4):

% uv run pytest -m integration -- --log-cli-level=DEBUG

test - urls

https://storage.googleapis.com/langrenn-sprint/photos/20240309_Ragde_lang.mp4 http://10.0.0.27:8080/video

slette images og containere

docker system prune -a --volumes

Push to github docker registry manually (CLI)

docker compose build docker login ghcr.io -u github password: Use a generated access token from GitHub (https://github.com/settings/tokens/1878556677) docker tag ghcr.io/langrenn-sprint/video-service:test ghcr.io/langrenn-sprint/video-service:latest docker push ghcr.io/langrenn-sprint/video-service:latest

Start service in google cloud run (only works in DETECT mode and with cloud_storage)

Create artifact registry and grant access to your service account

gcloud artifacts repositories create docker-repository
--repository-format=docker
--location="europe-north1"
--description="Docker repository"

gcloud projects add-iam-policy-binding [PROJECT_ID]
--member="serviceAccount:[SERVICE_ACCOUNT_EMAIL]"
--role="roles/artifactregistry.writer"

Upload Container

This is handled by github_actions - see .github/workflows/deploy_google.yml

Start Cloud Run instance

• Go to Worker Pools and Click Deploy Container
• Select "Depoloy one revision" and paste container image uri (or select from the list)
• Configure scaling (recommendation: ?)
• Paste environment variables from this file - remember to modify passwords!
• Environment - make sure MODE = DETECT
• Click create!