Skip to content
A vision-kinematic fusion approach for articulated tool tracking
C++ C CMake
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
config
data
include
model
src
CMakeLists.txt
LICENSE
README.md

README.md

daVinciToolTrack

A vision-kinematic fusion approach for articulated tool tracking. Example video results can be found from https://youtu.be/oqw_9Xp_qsw.

If you use the code, please cite the following paper:

https://arxiv.org/pdf/1605.03483.pdf

@inproceedings{ye2016real,
  title={Real-time 3D Tracking of Articulated Tools for Robotic Surgery},
  author={Ye, Menglong and Zhang, Lin and Giannarou, Stamatia and Yang, Guang-Zhong},
  booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
  pages={386--394},
  year={2016},
  organization={Springer}
 }

Prerequisites

QT 5.5+

OpenCV 2.49+.

VTK 6.30+

Eigen (included)

The CAD model files can be downloaded from http://research.intusurg.com/dvrkwiki/index.php?title=DVRK:Topics:InstrumentModels. But you may need to contact Intuitive Surgical (research@intusurg.com) for access.

License

This code is distributed under BSD License.

Notes

  1. The provided implementation tracks up to two articulated da Vinci tools.

  2. A modern CPU processor should enable real-time processing speed of this implementation.

  3. You can comment "#define MULTITHREADING" in "simulator.cpp" if you don't need parallel simulation and processing.

  4. Although a running example is given for offline processing, it is straightforward to combine it with kinematics retrieved from dVRK or Intuitive API for live tracking.

  5. This implementation has been tested both in Windows and Ubuntu.

You can’t perform that action at this time.