This is the project repository for Project No. 1 Extended Kalman Filter, part of Term 2 Sensor Fusion of Udacity Self-Driving Car Nanodegree program, submitted by Albert Killer in April 2017.
An Extended Kalman Filter (EKF) was implemented in C++ to process simulated LIDAR and RADAR measurements provided by Udacity. In doing so the position and velocity of an object, moving around a vehicle equiped with those sensors, can be detected with high accuracy:
Accuracy - RMSE:
px: 0.0972256
py: 0.0853761
vx: 0.4508550
vy: 0.4395880
I use a Kalman Filter for LIDAR data. In order to handle the nonlinear measurement function h(x) of RADAR data I apply an Extended Kalmand Filter (EKF). The EKF uses a linear approximation of h(x) and applies First Order Taylor Expansion in order to retrieve a Gaussian distribution again.
Want to know more about Extended Kalman Filters? Have a look at this one: https://en.wikipedia.org/wiki/Extended_Kalman_filter