CV - Outcognito

The basic, bare-bones set-up

Team Members

Rohan Arya Varma (ID: 598) Sujay R Subramanian (ID: 599) Tan Chin Khang (ID: 585)

Inspiration

There are many ways to secure transactions and information exchange in the virtual world, however your data could be easily compromised by a person just standing right behind you.

What it does

After detecting motion via a remote camera, a user-defined response is automatically executed in the display of interest, which can be one of the following:

Window Minimization
Screen Brightness Reduction
Opening Random Resource Site

An alert popup is also created, which upon confirmation will restore the original display.

How we built it

We used a Raspberry Pi 3 Model B+ with a Camera module to map the immediate surroundings of the user or entry point and detect motion. Using SSH, byte-sized data is transmitted over to the user's computer, which when received, will proceed to the action, which was selected via a GUI during the initialization process.

Challenges we ran into

Connection to Wi-Fi and Stability Issues
Efficiency in Image Processing
Visual 'Noise'
Data Management in Server
Difficult finding HWND

Accomplishments that we're proud of

Establishing a working communication protocol with auto-reconnect capabilities.
Optimization of Real-Time Image Capture and Processing for Raspberry Pi
Successful coding of multiple execution options

What we learned

Networking and Interfacing
Multiple libraries we had little to no experience with (subprocess, tkinter, ctypes, picamera, socket)
Interacting with computer display using different libraries
Technicalities regarding computer display (E.g. HWND, Use of Powershell)