Skip to content

Latest commit

 

History

History
96 lines (66 loc) · 2.77 KB

README.md

File metadata and controls

96 lines (66 loc) · 2.77 KB

Motion Detection

Build your own low cost security cam!

This is the backend, which is based on the python raspberry pi camera module interface picamera, for the coming soon iOS-App.

Please feel free to change and improve the code, this is my first python project.

Features

  • records in 1080p with low cpu usage
  • records also a few seconds before the motion was detected
  • creates a thumbnail of the event as preview for the video
  • define areas which don't need to checked (e.g. a tree in the wind)
  • optional nightmode with lower fps but much brighter videos
  • live view

Requirements

  • Raspberry Pi
  • Camera Module

Installation

Instructions for an installation on Raspbian Jessie Lite

  • Install dependencies
sudo apt-get install python-imaging gpac libav-tools imagemagick git python-dev python-pip
sudo pip install pyephem numpy picamera
  • Clone this repository
sudo git clone https://github.com/dneunkirch/motiondetection.git /etc/motiondetection
  • Create a symlink for the init script
sudo ln -s /etc/motiondetection/scripts/startup.sh /etc/init.d/motiondetection
  • Register the init script
sudo update-rc.d motiondetection defaults

Now you're able to start/stop the motion detection with the command sudo service motiondetection {start|stop}. The motion detection starts also automatically at startup.

  • Configuration-File

/etc/motiondetection/python/config.ini

URLs

The port (default: 8080) and credentials (default: username, password) can be changed in the config file. You're able to add multiple users in the users section or even turn the basic authorization off.

  • Live MJPEG stream /live or /live.mjpeg

  • Current live picture /live.jpg

  • Exclude areas from motion detection /blacklist.html

  • Delete an event /delete?file=filename_of_video.mp4

    A GET request to this endpoint deletes the given video including his preview image.

  • Captured Videos /events

    Returns an array of all captured videos as following object:

    {
  "date": "yyyy-MM-dd_HH-mm-ss",    // date of video
  "video": "/events/video.mp4",     // path to video
  "poster": "/events/preview.jpg",  // path to preview image
  "size": 1000000,                  // size of video in bytes
  "duration": 10                    // duration of video in seconds
}

For testing and debugging

  • Fake Motion /force_motion

  • Stop Faking Motion /stop_force_motion

  • Activate Daymode /daymode

  • Activate Nightmode /nightmode

    Works only if night_mode_allowed in the config is True

Todo's

  • submit iOS-App to App-Store
  • create a web-App
  • write tests