We provide code and executables for our 3D scene reconstruction system. The code and the executables are released under the MIT license. In general, you can use this for any purpose, including commercial applications, with proper attribution. We are releasing this system in hope that it will be useful in many settings. If you do something interesting based on our work, please acknowledge it. We'll be happy to know about various applications, feel free to send us an email and tell us about your work. The System Tutorial page will help you get started quickly. Please also let us know if you think our system can be improved. In particular, we are interested in effective geometric registration algorithms: see the Registration Evaluation page for more details.

Feel free to contact the authors: Qianyi Zhou (, Sungjoon Choi (, Vladlen Koltun (


The techniques that were used to create this system are presented in the following papers:

Robust Reconstruction of Indoor Scenes, CVPR 2015
Sungjoon Choi, Qian-Yi Zhou, and Vladlen Koltun
Simultaneous Localization and Calibration: Self-Calibration of Consumer Depth Cameras, CVPR 2014
Qian-Yi Zhou and Vladlen Koltun
Elastic Fragments for Dense Scene Reconstruction, ICCV 2013
Qian-Yi Zhou, Stephen Miller, and Vladlen Koltun
Dense Scene Reconstruction with Points of Interest, SIGGRAPH 2013
Qian-Yi Zhou and Vladlen Koltun

Please cite them properly if you use our code, system, or data. Specifically, if you use the reconstruction system, geometric registration evaluation setup, or the Augumented ICL-NUIM Dataset, cite our CVPR 2015 paper (Robust Reconstruction of Indoor Scenes). If you use the FragmentOptimizer routine to improve reconstruction quality with nonrigid registration, cite our ICCV 2013 paper (Elastic Fragments for Dense Scene Reconstruction) and/or our CVPR 2014 paper (Simultaneous Localization and Calibration: Self-Calibration of Consumer Depth Cameras). If you use data from our old 3D scene dataset, cite our SIGGRAPH 2013 paper (Dense Scene Reconstruction with Points of Interest).


The code is hosted on GitHub:

The compilation environment is Windows 7/8/8.1 64bit OS + Visual Studio 2010. The current codebase has many dependencies including PCL. You will need to compile PCL with VS2010 and only then compile our system. You will need to fiddle around a bit to make it work. We are not happy with the current compatibility issues. We are working on a new code release that will not depend on external libraries as much and will be much easier to compile. Stay tuned.


See the System Tutorial page for instructions on setting up the system and using it to reconstruct scenes.

See the Parameter Configuration page to learn more about the configuration of the system.