Skip to content
Branch: master
Go to file

Latest commit


Failed to load latest commit information.
Latest commit message
Commit time

============ Introduction

This repository is the code for the paper titled:

Modern MAP inference methods for accurate and faster occupancy grid mapping on higher order factor graphs by V. Dhiman and A. Kundu and F. Dellaert and J. J. Corso

Please refer to the paper for more details.

====================== Repeatable experiments


You will need

  1. Boost 1.48 sudo apt-add-repository ppa:jkeiren/ppa sudo apt-get update sudo apt-get install libboost*1.48-dev

  2. OpenCV 2.4 Install from source

  3. GTSAM 2.3 Install from source

  4. Gtest

  5. You can optionally install player and stage to generate simulated data again, which is enclosed.


  1. All the development and testing has been done on Ubuntu 12.04.

  2. Use standard cmake procedure for building binaries mkdir build cd build cmake .. make

  3. All executable binaries are created in "bin" folder in the root directory.

  4. The data required for the paper is available by running the python script in Data/ directory.

    cd Data/; python

Convergence Experiments

You can run the convergence experiment on "cave" dataset. The following scripts will run five algorithms on cave dataset:

python scripts/ scripts/

Similarly, for other datasets use the following scripts:

python scripts/ scripts/
python scripts/ scripts/

These scripts will save qualitative results in the respective data directories as *.png files. These scripts will save *-plot-time-energy.npy files in the respective data directories. This data can be plotted by using the following scripts:

python scripts/ scripts/
python scripts/ scripts/
python scripts/ scripts/

Step size comparison for dual decomposition

The following scripts will run and plot the dual decompositon

python scripts/ scripts/
python scripts/ scripts/

====================== Acknowlegements

Vikas Dhiman and Jason J. Corso were partially supported by FHWA DTFH61-07-H-00023. Jason J. Corso was partially supported by NSF CAREER IIS-0845282 and ARO YIP W911NF-11-1-0090. Frank Dellaert and Abhijit Kundu were partially supported by an ARO MURI W911NF-11-1-0046. We thank Stan Birchfield and Brian Peasley for discussions and early efforts in this work.


Modern MAP algorithms for Occupancy Grid Mapping




No releases published
You can’t perform that action at this time.