The algorithm was published in Dec. 2018 at SIGGRAPH Asia. Please refer to our project website (https://vccimaging.org/Publications/Smith2018UAVPathPlanning/) for the paper and further information.
The code and the build system have the following prerequisites:
- git
- make
- gcc (>= 5.4.0)
- cuda (>= 8.0.0)
- libglfw, libGL, libgomp, libpng, libjpg, libtiff
Furthermore the code depends on the following projects, some with adaptions so please clone and build the respective branches:
- Premake https://premake.github.io/
- Multi-View Environment https://www.gcc.tu-darmstadt.de/home/proj/mve
- MVS-Texturing https://www.gcc.tu-darmstadt.de/home/proj/texrecon
- Eigen http://eigen.tuxfamily.org
Starting any application without parameters will print a description and an explanation of the parameters.
git clone https://github.com/nmoehrle/uavmvs.git
cd uavmvs
premake5 gmake
cd build
make
(ormake -j
for parallel compilation)
Our software is licensed under the BSD 3-Clause license, for more details see the LICENSE.txt file.
If you use our capture planning code for research purposes, please cite our paper:
@inproceedings{Smith2018Aerial,
title={Aerial Path Planning for Urban Scene Reconstruction:
a Continuous Optimization Method and Benchmark},
author={Smith, Neil and Moehrle, Nils and Goesele, Michael and Heidrich, Wolfgang},
year={2018},
publisher={ACM}
}
If you have trouble compiling or using this software, if you found a bug or if you have an important feature request, please use the issue tracker of github: https://github.com/nmoehrle/uavmvs