Decided to get more serious into learning back coding. Technically I’ve been coding on and off with Python for internships or personal projects, but idk I always preferred C and C++ cause embedded programming is just more fun. Started doing Hackerrank just to get back in the swing of things.

Figured out why the Kalman Filter was not working- it’s good for tracking, but not so much as filtering noise. It can do that, but if there’s a sudden new object introduced it slowly increases thinking the two objects are the same.

So in this case it might help looking into different algorithms.

After working with both now, I can safely say STM32>>Arduino imo. Like sure, Arduino has a ton of libraries, a simpler IDE, tons of peripherals, tutorials etc. But idk, working with STM32 just feels better for me.

Looks like 1D Kalman Filter isn’t as good at filtering noise compared to when it’s working with a Linear model, plus you can see the lag from it in real-time. The raw data approach can still work thanks to the quality of an expensive sensor lol, just not as cool.

I want to add more advanced versions of it to the current code, but for the time crunch a vanilla Kalman Filter should be enough for now. Still it’s just not as cool as saying you added the Unscented Kalman Filter.

Really need to learn the Pogo techniques for Zeta. Idk if they’re even optimal, it just looks funny to pull off

Watching high level Zeta play is insanely fun! But also reminds me I only scratched the surface of all the stuff that’s possible