These are the dependencies used for project owl:
- flask
- cv2
- numpy
- telegram
Install these dependencies by running this command:
pip3 install -r requirements.txt
In order to start these services:
- Motion detection
- Telegram bot
- Alarm System
- Web Server
Simply run this command within the root directory of project owl in the raspberry pi:
python3 server.py
Apart from software, here is the hardware infrastructure:
And here is how it looks like after setting everythin up:
This section will guide you through the functions of this IoT device.
- Motion Detection
- Detects motion through image processing using computer vision 2 (cv2) in python.
- Main source of information for other services
- Automatically captures images when a motion is detected
- Telegram bot
- Allows users to communicate with the IoT device
- Ability to sound alarm at will
- A variety of commands for the user to use
- Alarm System
- Acts as a deterrant for any possible intrusions
- Alerts the owner if he/she is nearby
- Mainly uses motion detection as a trigger
- Web Server
- Allows the user to have a live view of the area
- Telegram
The invite link to the group chat with the bot is: https://t.me/+nwSxVbzIgu4xODM1
Here are the commands that the bot is listening out for when communicating with it.
- /start - Display this message
- /latest - Show the last picture taken
- /extractall - Send all images in an archive
- /beep - Beep the alarm
- /uptime - Shows the amount of time project owl has been running
- Webserver
-
navigate to the raspberrypi's ip address at port 8081 (e.g. http://raspIPaddr:8081)
-
default page shows the live stream of the camera
- desktop server.py
- This is the desktop version of server.py in which doesnt contain the RPi library
- The desktop version can be used the same way as the raspberry pi version with:
python "./desktop server.py"
- Email alerting has been changed to telegram reporting
- Reasons:
- Allows users to easily communicate with the bot
- Easier for the user to see telegram than to open an email app
- Telegram offers a free API for bots while emails are restricted to a sandbox environment
| Physical Part Used | Specifications |
|---|---|
| Circuit board PCB | 3cm x 7cm ZY2486A PCB |
| Raspberry pi | Model 4B 8gb RAM |
| Piezo buzzer | 3.3~24V buzzer |
| White LED | 3.3V LED light |
| Voltage booster | 5V to 12V booster |
| Top casing | 3D printed casing |
| Bottom casing | 3D printed casing |
| Customised camera casing | 3D printed casing |
| Miscellaneous parts | e.g. screws |
| Part of the project | Contribution |
|---|---|
| 3D printing + modelling of the case | Kai En |
| Planning and designing of the project | Kai En |
| Purchasing of all the necessary parts | Kai En |
| Programming of the telegram bot | Kai En |
| Programming of the motion detection | Kai En |
| Programming of the webserver | Kai En |
| Programming of the raspberry pi's GPIO pins | Kai En |
| Web page design | Kai En |
| File system infrastructure | Kai En |
| Circuit board creation | Kai En |
| Integration of all the parts | Kai En |
| Testing of the project | Kai En |
| Documentation | Kai En |
3D model of top casing: https://www.thingiverse.com/thing:922740
NOTE: Other items not listed here were originally created/designed