CVO/Adaptive CVO

• Continuous Direct Sparse Visual Odometry from RGB-D Images.
• Adaptive Continous Visual Odometry from RGB-D Images.

MATLAB Examples for CVO

For a toy example of registration using Kinect data in MATLAB run matlab/run_toy_example.m

Dependency

• ubuntu 16.04
• C++ 11 or higher
• Eigen3
• OpenCV 3.0.0
• PCL 1.4 (For saving pcd files)
• Intel C++ compiler (For better speed performance)
• Intel TBB
• Boost (for timing only)

Environment Setting

• Add the following two source command to your .bashrc, detailed instructions can be found here:

source opt/intel/mkl/bin/mklvars.sh intel64
source opt/intel/compilers_and_libraries/linux/bin/compilervars.sh intel64

Data

• Download TUM RGBD dataset from here.
• Generate the association files using data/rgbd_dataset/rgbd_benchmark_tools/assoc.sh.

cpp Code

To compile the cpp code, type the command below:

cd cpp/rkhs_se3_registration
mkdir build
cd build

If this is your first time compiling using intel compiler, set your cmake varaibles by the following command: (learn more here)

cmake .. -DCMAKE_C_COMPILER=icc -DCMAKE_CXX_COMPILER=icpc
make

Next time, you only need to do the following:

cmake ..
make

Then executable files cvo and adaptive_cvo will be generated in build.

To run cvo code:

./cvo $path_to_data $tum_sequence_number(1 for fr1, 2 for fr2, 3 for fr3)

A txt file containing the trajectory, cvo_poses_qt.txt, will be generated in your data folder.

To run adaptive cvo code:

./adaptive_cvo $path_to_data $tum_sequence_number(1 for fr1, 2 for fr2, 3 for fr3)

A txt file containing the trajectory, acvo_poses_qt.txt, will be generated in your data folder.

An example of how to run the code is in the script folder.

Citations

• Maani Ghaffari, William Clark, Anthony Bloch, Ryan M. Eustice, and Jessy W. Grizzle. "Continuous Direct Sparse Visual Odometry from RGB-D Images," in Proceedings of Robotics: Science and Systems, Freiburg, Germany, June 2019. https://arxiv.org/abs/1904.02266

@INPROCEEDINGS{MGhaffari-RSS-19, 
    AUTHOR    = {Maani Ghaffari AND William Clark AND Anthony Bloch AND Ryan M. Eustice AND Jessy W. Grizzle}, 
    TITLE     = {Continuous Direct Sparse Visual Odometry from RGB-D Images}, 
    BOOKTITLE = {Proceedings of Robotics: Science and Systems}, 
    YEAR      = {2019}, 
    ADDRESS   = {Freiburg, Germany}, 
    MONTH     = {June} 
} 

• Tzu-Yuan Lin, William Clark, Ryan M. Eustice, Jessy W. Grizzle, Anthony Bloch, and Maani Ghaffari. "Adaptive Continuous Visual Odometry from RGB-D Images." arXiv preprint arXiv:1910.00713, 2019. https://arxiv.org/abs/1910.00713

@article{lin2019adaptive,
  title={Adaptive Continuous Visual Odometry from RGB-D Images},
  author={Lin, Tzu-Yuan and Clark, William and Eustice, Ryan M and Grizzle, Jessy W and Bloch, Anthony and Ghaffari, Maani},
  journal={arXiv preprint arXiv:1910.00713},
  year={2019}
}