Extended Kalman Filter

This Project is first project of second term of Udacity Self-Driving Car Nanodegree program. The goal is to apply Extended Kalman Filter to data from LIDAR and RADAR sensors by C++.

Content

• scr contains source code:
  • main.cpp - loads data, calls a function to run the Kalman filter, calls a function to calculate RMSE
  • data_source.cpp - load input data into measurement and ground truth vectors
  • fusion_ekf.cpp - initializes the filter, calls the predict function, calls the update function
  • kalman_filter.cpp- defines the predict function, the update function for lidar, and the update function for radar
  • tools.cpp - a function to calculate RMSE and the Jacobian matrix
• data a directory with two input files, provided by Udacity
• result a directory with output files, log files and output charts
• docs a directory with files formats description

Results

In order to see how sensor data impact accuracy of Kalman Filter, I run the algorithm on two data input (porvided by Udacity) with three calculations:

1. Lidar and Radar measurements are considered
2. Only Lidar measurements are considered
3. Only Radar measurements are considered

Here is the RMSEs comparion of expected and above calculations:

Data 1

RMSE	px	py	vx	vy
Threshold	0.08	0.08	0.60	0.60
Radar + Lidar	0.0655724	0.0624793	0.534247	0.54817
Lidar	0.0681865	0.0607532	0.625587	0.570902
Radar	0.10121	0.0823387	0.601316	0.581942

Radar and Lidar measurements are considered

Only Lidar measurements are considered

Only Radar measurements are considered

Data 2

RMSE	px	py	vx	vy
Threshold	0.20	0.20	0.50	0.85
Radar + Lidar	0.185609	0.189915	0.473388	0.776187
Lidar	0.217995	0.19416	0.93745	0.800829
Radar	2.80105	2.67625	3.83749	4.39577

Radar and Lidar measurements are considered

Only Lidar measurements are considered

Only Radar measurements are considered

The results were visualized with Sensor Fusion utilities.

Lessons Learned

• Considering both sensor data (LIDAR and RADAR) give much better accuracy. You can see in RSME table that when we are considering both sensor data, we have acceptable RMSE.
• Lidar sensor give more accurate measurement than radar

Build Instructions

1. Clone this repo.
2. Make a build directory: mkdir build && cd build
3. Compile: cmake ../src && make
   • On windows, you may need to run: cmake ../src -G "Unix Makefiles" && make
4. Run it: ./ExtendedKF path/to/input.txt path/to/output.txt. You can find some sample inputs in 'data/'.
   • eg. ./ExtendedKF ../data/sample-laser-radar-measurement-data-1.txt output.txt true|false true|false