Inspiration

As a EE student, I work with resistors a lot, but i never remember what the colours stand for. Additionally, the solutions on google are all based on manual input, which is hard as resistors are tiny and also the colours kind of bleed into each other

What it does it do

It takes video input and runs it through an OpenCV object identifier, followed by filtering to determine the colour readings of the resistor.

How I built it

built with a simple html front end sending and displaying video, with a flask server backend

Challenges we ran into

this initially started off as a purely frontend project, with the user doing the processing on their end using opencv.js, however opencv.js used more then 4gb of memory space, exceeding the heap limit, so i had to pivot to another solution. integration of the server codebase with the OpenCV model was also challenging, as I had to translate the opencv2 code into opencv3.

Accomplishments that I'm proud of

That it even works in the first place, and can at least recognize resistors, while giving a better then human guess of the value.(what was your guess of the resistor in the thumbnail?)

What I learned

how fast coding is when you don't have to worry about MOSS also somewhat more comfortable with flask, react and also openCV

What's next for resistor picture

-change the html to only submit pictures as video processing poses an insane workload

-Use a custom lightweight version of opencv.js through emscripten

  • port the project to javascript so it is lightweight and mobile friendly

Built With

Share this project:

Updates