Description
I recently got an OpenMV, which is an open source machine vison platform. It is a 640x480 TTL camera slapped onto a board with and STM32 processor. There is also an IDE with a bunch of examples to get you started. The one thing I am on the fence about is the default laguage for the IDE, which is MicroPython. I'm split because on the one hand I love python and it is quick to get started on cool vision project, but on the other it's still python. What you gain in ease of use you lose in performance. Luckily the entire project is on GitHub and all the algorithms are implemented in C. However before I dive back into C I wanted to give MicroPython a try. I wanted to see the performance difference from implementing the same project in both C and MicroPython.
So that brings me to the Pixelizer. The idea is pretty simple, take the camera image data and convert it to a display on a 84x48 monochrome screen from a late 90's cell phone. I want the conversion to happen in real-time (25fps minimum) and I want to be able to take and save pictures. I'll implement the project both in MicroPython and C using the same (mostly) conversion algorithm.