Extended-Kalman-Flter

Implementation of EKF with sensor fusion to track the position of a car in real time, using data streams from RADAR and LIDAR.

---

Build Instruction

1. Clone this repo and cd into it.
2. mkdir build && cd build
3. cmake ..
4. make
5. Run : ./ExtendedKF

Once the program starts running, the following message can be seen : Listening to port 4567. Note: This project requires the Udacity open source simulator : Udacity term 2 sim.

---

Dependencies

• cmake >= 3.5

  • All OSes: click here for installation instructions

• make >= 4.1 (Linux, Mac), 3.81 (Windows)

  • Linux: make is installed by default on most Linux distros
  • Mac: install Xcode command line tools to get make
  • Windows: Click here for installation instructions

• gcc/g++ >= 5.4

  • Linux: gcc / g++ is installed by default on most Linux distros
  • Mac: same deal as make - install Xcode command line tools
  • Windows: recommend using MinGW

  ---

  Result

  The make file runs without any errors. The RMSE for dataset 1 is as follows :

  Input	MSE
  px	0.0974
  py	0.0855
  vx	0.4517
  vy	0.4404

  The RMSE for dataset 2 is as follows :

  Input	MSE
  px	0.0726
  py	0.0965
  vx	0.4219
  vy	0.4937