ShudaLi / rgbd_mapping_and_relocalisation

This is an open source C++ implementation of the paper RGB-D Relocalisation Using Pairwise Geometry and Concise Key Point Sets. Li, S., & Calway, A. ICRA, 2015.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool


Dense RGB-D Mapping and Relocalisation


Introduction

This is an open source C++ implementation of [1]. It provides scene mapping and fast and robust wide baseline relocalisation without the need for training. The code has been tested on both Ubuntu and Windows platforms.

[1] RGB-D Relocalisation Using Pairwise Geometry and Concise Key Point Sets. Li, S., & Calway, A. ICRA, 2015. [2] Absolute pose estimation using multiple forms of correspondences from RGB-D frames. Li, S., & Calway, A. ICRA, 2016.

Please cite above publication if you use this software in your own work.

The repository is maintained by Shuda Li (lishuda1980@gmail.com). Feel free to contact me, if you have any questions or suggestions.


Functional keys

'7' switch on the dense surface reconstructed.

'0', pressing 0 key once aligns the viewing position of up left sub window with current camera pose only. pressing 0 key one more time locks the viewing position with the camera location and pressing 0 again release the lock.

'F5', switch on/off the recorded camera path (red/blue curves)

'M', convert the TSDF surface representation into a triangle mesh, to display it, press '7'

More key settings can be found in Viewer::keyPressEvent() in the ./live_mapping_and_relocalisation/MultiViewer.cpp


License

The source code is released under the MIT license. In short, you can do anything with the code for any purposes. For details please view the license file in the source codes.


Dependences

  • OpenCV 3.0 (BSD license)
  • Eigen (GNU free software)
  • Sopuhs (GNU free software)
  • OpenNI 2 (Apache license)
  • GLEW (Free commercial use)
  • NIFTI (Public domain)
  • ZLIB (Free commercial use)
  • CUDA

Compilation


Demo Video and talk


Project page

About

This is an open source C++ implementation of the paper RGB-D Relocalisation Using Pairwise Geometry and Concise Key Point Sets. Li, S., & Calway, A. ICRA, 2015.


Languages

Language:C++ 71.3%Language:C 19.3%Language:Cuda 3.1%Language:Assembly 1.8%Language:Ada 1.1%Language:Pascal 0.9%Language:C# 0.7%Language:Makefile 0.5%Language:HTML 0.4%Language:DIGITAL Command Language 0.4%Language:CMake 0.2%Language:Roff 0.1%Language:CLIPS 0.1%Language:Perl 0.1%Language:SAS 0.0%Language:Module Management System 0.0%Language:Objective-C 0.0%Language:M4 0.0%Language:Batchfile 0.0%