DE-TECHT
An efficient and intelligent detection system.
Created on 12th July 2020
•
DE-TECHT
An efficient and intelligent detection system.
The problem DE-TECHT solves
Three main applications of this project are:
-
Developing a home appliance automation system based on the detection of faces and movements in the room. (Improved lockdown experience)
-
Developing a smart navigation system by analyzing road traffic based on the detection of moving vehicles. Detection data obtained could then be integrated with google maps to give users real-time, constantly updated traffic information at any time of the day. ( Improved navigation)
-
Developing an automated sanitizing booth, which will be a small gateway where a camera with our program is installed.
Mechanized sanitizers integrated to the camera could then spray liquid based on movement detected by the camera. (Improving health and fitness after the pandemic)
Besides these three applications, our project can also be used as a simple surveillance system.
Challenges we ran into
Webcam camera quality was not very good in my computer.
Sometimes small objects were detected as faces , though they were not faces.
The motion detector went awry sometimes when the first frames captured were not perfectly still.
We could not do much about the webcam quality, but we used certain permutations and combinations of function parameters to detect only faces as precisely as possible.
We started the motion detector in a perfectly silent room with now winds and no movements to get a very still first capture frame.
Cars moving with a very high speed were sometimes left undetected by our car detector. However, we tried to minimize this issue by referring to the OpenCV documentation and trying out using various other functions and parameters.
With a little bit of more time, we would have been able to build a slightly more accurate and efficient detecting system. Hopefully we have done good for our first hackathon!