Skip to content
/ mcl-cpp Public

Monte Carlo Localization using Particle Filtering

Notifications You must be signed in to change notification settings

mlsdpk/mcl-cpp

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

3 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Monte Carlo Localization using Particle Filtering

This repository represents the localization of a robot using the popular Monte Carlo Localization Algorithm. It is also called Particle Filter Localization Algorithm since it uses particles to localize the robot. The algorithm is implemented in C++.

alt text

This project is part of a Udacity Robotics Software Engineer NanoDegree. You can find more information of the NanoDegree here.

Compiling the program

Since virtual environment is used, you don't need to install any requirements for this. You only need g++ compiler to run the C++ source code.

In order to run the source code, you need to edit the shell script inside the MCL directory. Edit the {path-to} line in the script to your path of this repository.

g++ main.cpp -std=c++11 -I /{path-to}/mcl-cpp/env/include/python2.7 -lpython2.7 -I /{path-to}/mcl-cpp/env/lib/python2.7/site-packages/numpy/core/include
./a.out

After that, activate the virtual environment and run the following command to execute the program:

source env/bin/activate
./run.sh

You can see that the code is executing and the algorithm is run for 50 steps. The following figure shows the last step of the algorithm and it shows that robot is well localized with some error amounts of particle samples around the robot.

alt text

About

Monte Carlo Localization using Particle Filtering

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published