Kalman Filter in C++
Tracking a thrown ball through noise
- 90%lower latency than single-threaded multiplication
A 2D projectile tracking simulator that estimates position and velocity from noisy sensor data in real time, using a discrete Kalman filter on a custom templated C++ matrix implementation.
I added move semantics and multithreaded multiplication to the matrix class, cutting latency 90% compared with single-threaded execution.