This project use Extended Kalman Filter (EKF) to fuse lidar and radar measurement to track the object's position and velocity.

To run this program, use Term 2 Simulator and uWebSocketIO, follow the steps below:

1. cd build
2. cmake .. && make
3. ./ExtendedKF
4. open simulator and select dataset

• Dataset 1

• Dataset 2

This program is built on the starter code provided by Udacity CarND team.

Some key modifications are listed below:

• Initialize state and covariance matrix using first measurement;
• Update F and Q by time elapsed;
• Predict state;
• Update measurement for lidar and radar;
  • Use Hj for radar update;
  • Add a function 'convert2measure' in Tools, to convert state (px, py, vx, vy) into radar measure (rho, phi, rho dot);
  • normalize phi in y to make sure it is within -pi to pi;
• Calculate RMSE;
• Handle small values (close to 0) in denominator.