Self-Hosted Home Security Camera System

A privacy-focused, intelligent security camera system.

Features:

• Multi-camera support w/ minimal configuration. Supports USB cameras and the Raspberry Pi camera module.
• Motion detection that automatically saves videos and lets you view them in the web app.
• Encrypted in transit, both from the cameras to the server and the server to your browser.
• Integrated authentication
• Self-Hosted
• Free and Open Source (GPLv3)

Example screenshots

Multiple live video feeds on a single UI:

Replay of recorded videos triggered by motion detection:

Setup

Sec-cam-server is designed to be deployed flexibly. There are 2 core components:

Camera:

A device capable of capturing video and streaming to the server. Requirements:

• Must be able to run Python
• Must have a device capable of capturing video (e.g. a USB camera, webcam, or raspberry pi camera module)
• Must be able to communicate to the server over a network. During setup, you will point the camera at the server's hostname or IP address.

Server:

The server components aggregates feeds from the cameras, performs motion detection, and makes video available in a web UI via an API. The server can be a standalone host, or can run on one of your cameras (e.g. a Raspberry Pi). Requirements:

• Must be able to run Docker.
• Must be accessible to your camera(s) over a network.

Example deployment configurations:

Example 1: A standalone server, with the hostname sec-cam-server.local, and 2 cameras, with hostnames camera1.local and camera2.local.

1. On sec-cam-server.local, follow the instructions in "Setting up the server". Replace <server-hostname:server-port> with sec-cam-server.local:8443.
2. On each camera, follow the instructions in "Adding a camera". In Step 3, run.sh should be updated to contain --server_url sec-cam-server.local.
3. The web UI with camera feeds will be available at https://sec-cam-server.local:8443

Example 2: 2 cameras with hostnames camera1.local and camera2.local, with the "server" running camera1.local

1. On camera1.local, follow the steps to install the server.
2. On camera1.local, follow the steps to install the in "Adding a camera". In Step 3, run.sh should be updated to contain --server_url localhost.
3. On camera2.local, follow the steps to install the in "Adding a camera". In Step 3, run.sh should be updated to contain --server_url camera1.local.
4. The web UI with camera feeds will be available at https://camera1.local:8443

Setting up the server

Docker:

1. Install Docker following the instructions on their website.
2. Clone this repository
3. Generate SSL certificates: ./create-certs.sh. Alternatively, you may place your own certs in the certs dir
4. Pull and run the docker containers:

   docker-compose pull
   API_URL=<server-hostname:server-port> docker-compose up -d

For example, if the server was running on the host sec-cam-server and port 8443 (the default), you should use API_URL=sec-cam-server:8443. 5. You should now be able to view the UI at https://<server-hostname>:8443. 6. Until a user is created, you will be automatically redirected to

Configuration:

Adding multiple users:

By default, user registration will be disabled once an initial user is created. To enable registration to add additional users, update the environment variable ENABLE_REGISTRATION under the server service:

- ENABLE_REGISTRATION=1

When you're done adding users, you should re-set this value to 0 and restart the server.

Adding a camera

Installation:

NOTE: These instructions assume you are deploying to a Debian-based OS.

1. Install the required packages and dependencies: sudo apt install python3-opencv libatlas-base-dev a. If you are installing on a Raspberry Pi using a Pi Camera module, also run sudo apt install python3-picamera2.
2. Clone this repository.
3. Create a virtual environment, inheriting site packages: python -m venv .venv --system-site-packages. You can activate this virtual environment with source .venv/bin/activate.
4. Install the package: cd backend && python3 -m pip install .[streamer].
5. Update --server_url in run.sh to point at the host you deployed the server to.
6. You can install the camera software as a systemd service or run it manually. Either:
   1. Run the script to create the camera systemd service: cd scripts && ./create-streamer-service.sh
   2. Run the camera manually: ./run.sh.
7. In the Web UI, you should see live video from that camera.

Contributors

• @khlam for his help with Github Actions and building docker images