2. Introduction
We built a automated security camera
Our system can move and track people
This is an improvement over the conventional security camera because:
It will only record when there is movement, and when there is a face
When a face is detected, alerts will be sent to the user (E-mail, SMS) that there is
a person of interest
Ability to monitor a large area with one camera
3. Motivation
Security Issues
Our interest in robotics and actuation and image processing
To reduce the cost and energy reduction of video monitoring and surveillance
Creating a system that functions and processes information in real time
4. Applications
Conventional camera very inefficient (Memory, FOV, power)
Need only one camera to secure an area vs multiple
Remote monitoring
Home Security
5. Group Members
Bogdan
Emailing images and
sending SMS to user
Manual Control of
Camera
Power Supply for
System (120V AC and
DC Batteries
Design of Stand Base
Messan
Microcontroller
Programming
Servo Motor Control
Sensor Configuration
Stand Articulation
Design
Image Processing
Object Tracking
GUI Design
Camera Calibration
Design of Camera
Mounting Surface
Chris
6. Project Development
We used Plexiglas as the main building material due to its low cost, high
strength and durability
Rotating platform that will be supported by a
bearing to allow easy movement
Arm that supports the camera and moves it
vertically
Material Cost
Plexiglas ~ $80
Servo Motors ~ $60
L-Brackets/Bolts/Bearings ~ $20
Microcontroller/Leads/Batteries/Sensors…etc ~ $60
Camera ~ $20
Total: ~$240
9. Features
SMS/Email
The moment a face is detected, an SMS is sent to the user alerting them to check
their email for an image of a possible intruder
Auxiliary Power
There are two onboard battery packs that allow for the device to still operate even
if external power is lost
Bluetooth
Communication between the Microcontroller and the computer can be done via
USB serial or Bluetooth to allow for wireless control
Manual Control
Control of the system can be done via Xbox controller or Bluetooth-enabled
smartphone
10. Project Development
GUI and tracking algorithms coded in Python
Opencv, PySerial, PyGame, Tkinter
Communication between computer and
Microcontroller done through USB serial or
Bluetooth Serial
Various parameters and settings can be
changed in the GUI
11. Tracking
Tracking starts the moment a
face is detected
Notification is sent to the user
Servos move the camera so that
the face stays in the center of
the screen
https://youtu.be/AQ-ZPdWxF4g
12. Problems encountered
Instability (Python/Arduino)
Lack of documentation for packages (opencv)
Parts failure and replacement
Cracking Plexiglas
Structural rigidity
Face detection false positives
13. Improvements and Future Development
With additional funding:
More expensive servos allow for quicker, more accurate control
More powerful Microcontroller allows for lower latency and quicker computation
times
Additional paid packages allow for better face/body recognition
With additional time:
Person identification
Smaller design for better fitment in confined spaces
14. Conclusion
Overall our team was able to overcome many obstacles in the developmental
and design phase
We are able to meet and exceed all the objectives that we have set fourth in
our project proposal
We were able to experience team working, problem-solving and debugging in
real world engineering application