Skip to content

sair-lab/correlation_flow

Repository files navigation

Correlation Flow

Correlation Flow: Robust Optical Flow using Kernel Cross-Correlators

	Velocity Estimation in 3-D space $v_x, v_y, v_z, \omega_z$

This repo contains source codes for the following paper, which is accepted by ICRA-18:

Chen Wang *, Tete Ji *, Thien-Minh Nguyen, and Lihua Xie, "Correlation Flow: Robust Optical Flow Using Kernel Cross-Correlators", IEEE International Conference on Robotics and Automation (ICRA), 2018.

Platform

Codes have been tested on Ubuntu 16.04 with ROS kinetic.

Install Dependencies:

  1. Install FFT library: sudo apt-get install libfftw3-dev libfftw3-doc

Install Intel Math Kernel Library (MKL):

MKL is optional but highly recommended for computational efficiency. It improves the running speed a lot.
  1. Download MKL from Intel website
  2. Extract downloaded file tar -zxvf [name_of_downloaded_file]
  3. Go to extracted folder, give permission: sudo chmod +x install.sh
  4. Run installation ./install.sh
  5. Link library, add to .bashrc: source /opt/intel/bin/compilervars.sh intel64
  6. Try compile in ROS workspace

If this repo is useful for your projects, you may cite it as:

@inproceedings{wang2018correlation,
  title={{Correlation Flow: Robust Optical Flow using Kernel Cross-Correlators}},
  author={Wang, Chen and Ji, Tete and Nguyen, Thien-Minh and Xie, Lihua},
  booktitle={International Conference on Robotics and Automation (ICRA)},
  year={2018},
  organization={IEEE}
}

Correlation Flow is built on following work:

Chen Wang, Le Zhang, Lihua Xie, Junsong Yuan, Kernel Cross-Correlator, In AAAI Conference on Artificial Intelligence (AAAI-18), 2018 (PDF available here) (source codes available here)

Works using Correlation Flow

Thien-Minh Nguyen, Abdul Hanif Zaini, Chen Wang , Kexin Guo, and Lihua Xie, "Robust Target-relative Localization with Ultra-Wideband Ranging and Communication", IEEE International Conference on Robotics and Automation (ICRA 2018), 2018. (Video available here)

*The above work applys correlation flow to improve the performance of localization accuracy.