jingjieli / BundleFusion

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

BundleFusion

BundleFusion

You are free to use this code with proper attribution in non-commercial applications (Please see License.txt).

More information about this project can be found in our paper and project website.

Apr. 25, 2018 Update - Build on VS2017 with CUDA9.1 for nVidia K2000

A few things you may have to do in order to compile:

  • If you're running VS2017 with latest update, most likely you're using MSVC++ 14.14 _MSD_VER == 1914. In "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1\include\crt>type host_config.h", modify _MSC_VER > 1911 to 1914

  • Compile error "cannot include correct.h" etc. Install Windows Universal CRT. Use Visaul Studio installer, choose "modify" for this additional package.

  • Cannot find DX hearder even after installing DirectX SDK June 2010. In Project Property VC++ Include Directory, add "" between "{DXSDK_DIR}" and "Include"

  • Error related to "std::enable_if" "xtr1common" blah blah. Turns out toolset 14.14 doesn't work well with CUDA9.1. Roll back to 14.11 works. Article on having two minor version of toolset by Microsoft: https://blogs.msdn.microsoft.com/vcblog/2017/11/15/side-by-side-minor-version-msvc-toolsets-in-visual-studio-2017/

  • And here's a general guideline to build, which may or may not work for your system: https://bericht.neopostmodern.com/posts/artist-guide-to-bundlefusion

Installation

The code was developed under VS2013, and tested with a Structure Sensor.

Requirements:

  • DirectX SDK June 2010
  • NVIDIA CUDA 7.0
  • our research library mLib, a git submodule in external/mLib
  • mLib external libraries can be downloaded here

Optional:

  • Kinect SDK (2.0 and above)
  • Prime sense SDK

Citation:

@article{dai2017bundlefusion,
  title={BundleFusion: Real-time Globally Consistent 3D Reconstruction using On-the-fly Surface Re-integration},
  author={Dai, Angela and Nie{\ss}ner, Matthias and Zoll{\"o}fer, Michael and Izadi, Shahram and Theobalt, Christian},
  journal={ACM Transactions on Graphics 2017 (TOG)},
  year={2017}
}

Contact:

If you have any questions, please email Angela Dai at adai@cs.stanford.edu.

We are also looking for active participation in this open source effort making large-scale 3D scanning publically accessible. Please contact us :)

About

License:Other


Languages

Language:C++ 66.7%Language:C 26.8%Language:Cuda 5.7%Language:RPC 0.4%Language:Objective-C 0.3%Language:HLSL 0.1%Language:FLUX 0.0%Language:Batchfile 0.0%