/ros_object_fusion

Object fusion based on a simplified interpretation of Aeberhard (2017) fusion architecture.

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

This is a ROS1 workspace to the object fusion algorithm based on Aeberhard (2017) http://dx.doi.org/10.17877/DE290R-18029 https://eldorado.tu-dortmund.de/handle/2003/36011

This work is a simplified implementation of the proposed fusion architecture below. Aeberhard proposed Fusion Architecture(Aeberhard(2017))

Install dependencies:

pip install numpy==1.21.2 rospkg==1.3.0 scipy==1.7.1

Compile the workspace

catkin build

Execute the code with:

rosrun object_fusion fusion.py

Visualize the output data (fused object markers) in RVIZ with:

rosrun object_fusion visualization_fusion.py

The fusion input in this example uses two subscribers, but it is not limited to those listes, any sensor with the same ROS1 message is accepted. The subscriber can be found in the fusion.py file as seen below.

    camera_subscriber = rospy.Subscriber("/fusion/input/sensors/camera", Object_List, callback_handler.callback)
    radar_subscriber  = rospy.Subscriber("/fusion/input/sensors/radar" , Object_List, callback_handler.callback)

(TO DO:)In case you use this code, please cite it with:

@INPROCEEDINGS{10066615,
  author={Poledna, Yuri and Reway, Fabio and Drechsler, Maikol Funk and Huber, Werner and Icking, Christian and Ribeiro, Eduardo Parente},
  booktitle={2022 10th International Conference in Software Engineering Research and Innovation (CONISOFT)}, 
  title={An Open-Source High-Level Fusion Algorithm in ROS for Automated Driving Applications}, 
  year={2022},
  volume={},
  number={},
  pages={174-181},
  doi={10.1109/CONISOFT55708.2022.00031}}