My implementation of Motion-Project/motion in Docker, including some scripts for ease of event integration.

My work in here is under the MIT license, so you can do with it what you will, even comercially if you like- however please note that Motion-Project/motion appears to be under under GPL-2.0 license which will likely limit your ability to do that.

A running instance of motion-cctv consists of the following:

• A Docker container with a Supervisor entrypoint, managing:
  • nginx
  • motion
  • event_parser_loop.py
  • logrotate

• In the configs folder
  • motion.conf
    • a basic configuration that suits my use case
  • camera1.conf, camera2.conf, camera3.conf
    • configurations for my 3 x Hikvision DS-2CD1H41WD-IZ cameras
• In the res folder
  • config.py
    • HTML templates for event_parser.py
  • event_parser.py
    • parses pictures and movies from target_dir and builds events.html
  • event_parser_loop.py
    • runs event_parser.py on a 60-second loop
  • event_parser_test.py
    • unit tests for event_parser.py
  • index.html
    • landing page
  • motion-cctv.conf
    • supervisor configuration for all the processes
  • nginx.conf
    • nginx web server configuration

• Define your cameraN.conf files
• Adjust motion.conf to suit
• Download the repo
  • ./get.sh
• Build the image
  • ./build.sh
• Run the instance in the foreground (you'll likely need to adjust the volumes)
  • ./run.sh
• Browse to http://localhost and have a click around

If you want to run the instance in a more productionized way then run_in_background.sh shows a suitable example (but again, you'll need to adjust the volumes).