Crowd-SLAM

Crowd-SLAM is a visual SLAM system that is robust in crowded scenarios.

Demonstration video: https://www.youtube.com/watch?v=LeS8MEVaR2E

Paper: https://link.springer.com/article/10.1007/s10846-021-01414-1

License

Crowd-SLAM is released under a GPLv3 License.

If you use Crowd-SLAM in an academic work, please cite:

@article{soaresJINT2021,
  title={Crowd-{SLAM}: Visual {SLAM} Towards Crowded Environments using Object Detection},
  author={Soares, J. C. V., Gattass, M. and Meggiolaro, M. A.},
  journal={Journal of Intelligent & Robotic Systems},
  volume={102},
  number={50},
  doi = {https://doi.org/10.1007/s10846-021-01414-1},
  year={2021}
 }

Building Crowd-SLAM

• Install ORB-SLAM2 prerequisites: C++11 or C++0x Compiler, Pangolin, OpenCV and Eigen3 (https://github.com/raulmur/ORB_SLAM2).
• Clone the repository:

git clone https://github.com/virgolinosoares/Crowd-SLAM

• Execute:

cd Crowd-SLAM
chmod +x build.sh
./build.sh

We have tested the library in Ubuntu 18.04, with OpenCV 3.4.

RGB-D Example (TUM Dataset)

• Download a sequence from http://vision.in.tum.de/data/datasets/rgbd-dataset/download and uncompress it.

• Associate RGB images and depth images using the python script associate.py:

  python associate.py PATH_TO_SEQUENCE/rgb.txt PATH_TO_SEQUENCE/depth.txt > associations.txt
  

• Execute the following command. Change TUMX.yaml to TUM1.yaml,TUM2.yaml or TUM3.yaml for freiburg1, freiburg2 and freiburg3 sequences respectively. Change PATH_TO_SEQUENCE_FOLDER to the uncompressed sequence folder. Change ASSOCIATIONS_FILE to the path to the corresponding associations file.

  ./rgbd Vocabulary/ORBvoc.txt Examples/RGB-D/TUMX.yaml PATH_TO_SEQUENCE_FOLDER ASSOCIATIONS_FILE
  

Acknowledgements

Our code builds on ORB-SLAM2.