The project is designed as a end to end solution for a DIY Home Monitoring & Intruder Alert system. Besides offering a live video stream on any device (web responsive client), it also actively monitors for movement with the help of a PIR sensor.
If an Alarm is triggered, you get a SMS notification on your phone and the snapshots taken during the Alarm time span (customizable - default is 1 minute) are uploaded via FTP to your server.
Activation / Deactivation of the Alarm Mode can be done in 2 ways:
- from the Web Client user interface
- with a Button - for convenience reasons: it is faster than connecting from your phone / pc & toggling the Alert Mode checkbox
- there is a 10 seconds customizable delay which allows you to move out of the PIR sensor range
- a Led indicates the Alarm Mode enabled/disabled status
In order to avoid false positives from the PIR motion sensor, extra checks were added - a detection counter & detection interval. The Alarm gets triggered when the sensor detects movement 3 times in 5 seconds (both values configurable in code).
The project was developed using:
- Raspberry Pi - raspbian, brick button & led, Pir sensor
- Node.js - for the main application
- Mjpg_streamer - to generate the video stream
- Shell scripting - for easy application start (interactive & background)
- Htms/Css/Javascript + Bootstrap - the web client
this.Gpio = require('pi-gpio');
this.Hardware = { MotionSensor : 8, Led : 26, Button : 12 };
- Raspberry Pi
- I used Model B Revision 2 with Raspbian - any model should be Ok, just be careful with the Gpio configuration pin mappings, they can differ
- Generic USB webcam (compatible with Raspberry Pi & Raspbian)
- You can find a comprehensive list here http://elinux.org/RPi_USB_Webcams
- I used a very old 2MP one which seems to work out of the box with the generic drivers
- Led & Button
- PIR motion sensor
- The one I used is available here https://www.sparkfun.com/products/13285
- It normally connects to Analog Input (ex. on Arduino); however you can use it with Digital as well if you connect a 10K resistor between VCC & Signal
- To make things easier you can purchase this sensor https://www.adafruit.com/products/189 and skip the soldering part (+ this one has configurable sensitivity built-in, so you might be able to skip the one implemented in the code)
- express: ^4.12.3
- ftp: ^0.3.10
- http-auth: ^2.2.8
- ini: ^1.3.4
- pi-gpio: 0.0.7
- socket.io: ^1.3.5
- twilio: ^2.3.0
The dependencies you install with NPM:
npm install module --save
It is the basic application object, defined to be reusable in other projects Contains the basic server code, generic config file read/write operations, generic Init & Execute & Exit methods implementations
-
config.ini file
- default video quality & alert mode settings
- Twilio sms Api Sid, Token, To number, From number
- Ftp settings
-
Authentication (digest http authentication) - defaults are admin & password :)
- You can change them from the
htdigest
file (nice helper tool here http://websistent.com/tools/htdigest-generator-tool/ )
admin:Private:6982db7f1ddc36a0b47b5f8427dc3526
- You can change them from the
-
Web Client application
- Accessible from anywhere via port forwarding
- Available also on mobile (responsive web client)
-
Monitoring - gets video from Mjpg_streamer server and sends it to the connected app clients
-
Mjpg_streamer was used as server, but if you prefer another tool like ffmpeg, you can easily replace it because of the loose integration via the
start-webcam.sh
script
- Monitoring - via PIR sensor
- Alarm - Sms notification (implemented with the help of Twilio text messaging API - very cool service, offers great Trial account for development
- Alarm - Snapshots upload to server via Ftp
The client application was designed to be accessible on all platforms (pc / tablet / mobile).
By default the 480p at 25fps is enabled (initial settings are loaded from the config.ini
file)
My webcam is a low-end 5+ years old 2mp device, but for those of you with better webcams I also added 720p & 1080p
Video resolutions & fps can be configured from the /static/js/script.js
file
ui.quality480p.change(function(){ ConfigUpdateQuality("640x480",25); });
ui.quality720p.change(function(){ ConfigUpdateQuality("1280x720",15); });
ui.quality1080p.change(function(){ ConfigUpdateQuality("1920x1080",5); });
- initial state is loaded from the
config.ini
file - You can enable/disable monitoring from checkbox button in the UI
- The state of the Alert Mode is shown both in the UI (the checkbox) but also by the LED
- The physical Button can be also used to toggle the Alert Mode
- All state changes are sent to all connected clients
- If an Alarm is triggered, the UI checkbox button background will be changed to Red
The dropdown shows a list of all connected clients (connection timestamp & IP) that are currently viewing the video stream
start-app.sh
- You can start the application in 2 modes:
- Interactive (for dev / testing):
./start-app.sh
- Background:
./start-app.sh -background
- Interactive (for dev / testing):
#!/bin/bash
# application start in interactive or background mode
#arguments: [-background]
cd /home/pi/Desktop/rpiWorkspace/Node/HomeMonitoring/
if [ "$1" = "-background" ]; then
sudo nohup node ./App-home-monitoring.js &>log.txt &
else
sudo node ./App-home-monitoring.js
fi
start-webcam.sh
- Used by the application to enable/disable video streaming when clients are connected or when an Alarm is triggered by the PIR sensor.
#!/bin/bash
# webcam video stream
# arguments: [resolution] [port] [fps]
pkill mjpg_streamer
sudo nohup ./mjpg-streamer/mjpg_streamer -i "./mjpg-streamer/input_uvc.so -y -r $1 -f $3 -q 75" -o "./mjpg-streamer/output_http.so -n -p $2" &
TO DO
- Port the application to Windows 10 Iot on Raspberry Pi 2
- Support for uploading snapshots to cloud (OneDrive / Dropbox) when an Alarm is triggered
References
- Raspberry Pi https://www.raspberrypi.org/
- Node - https://nodejs.org/en/
- Mjpg_streamer http://sourceforge.net/projects/mjpg-streamer/
- SMS Api - Twilio - https://www.twilio.com/sms
- Bootstrap http://getbootstrap.com/
- App Webcam Icon - https://www.iconfinder.com/icons/71274/webcam_icon#size=128
Links
- Blog post about project Home Monitoring with Raspberry Pi & Node
- Hackster.io project page
- twitter @orosandrei