dongyu728 / carnd-ukf

CarND-Unscented-Kalman-Filter

Udacity Self-Driving Car Engineer Nanodegree Program, Term 2 ,Project 2 - Unscented Kalman Filter

Project Basics

This project is write by C++.

I used Unscented Kalman Filter to predict and update the position based on lidar and radar data.

This project involves the Term 2 Simulator.

Build instructions

I was build in Ubuntu 16.04:

1. Clone this repo.
2. Make a build directory: mkdir build && cd build
3. Compile: cmake .. && make
4. Run it: ./ExtendedKF

Results

There are two dataset.
px:x-position
py:y-position
vx:velocity in the x-direction
vy:velocity in the y-direction
MSE:mean squared error

The UKF accuracy was:

Dataset 1 : RMSE = [0.0724, 0.0822, 0.3425, 0.2304]

Dataset 2 : RMSE = [0.0882, 0.0713, 0.6632, 0.3073]

Test One

Test Two