mvs-texturing

nmoehrle更新放缓

从多视角立体图像给三维重建模型自动生成纹理贴图，解决重建网格缺少真实感表面外观的问题。

Algorithm to texture 3D reconstructions from multi-view stereo images

暂无版本1,112star371fork同步于 2026-11-07 07:30 UTC

纹理贴图三维重建多视角立体网格纹理 C++

对比 GitHub 官网 3d-reconstruction mvs texturing

怎么上手

由 README 提炼，执行前请确认

原文

git clone https://github.com/nmoehrle/mvs-texturing.git
cd mvs-texturing
mkdir build && cd build && cmake ..
make -j

硬件要求按 README 说明

README 未说明显存或 GPU 要求，仅列出编译依赖（cmake、gcc、libpng/libjpg/libtiff/libtbb），未提及 GPU 厂商或型号

下载：源码 zip·GitHub Releases（安装包）本站不镜像安装包

项目说明

由 README 自动翻译提炼

没人工点评？写下你的经验

给多视图立体重建出的三维模型自动生成纹理贴图的命令行工具

从多视角立体图像给三维重建模型自动生成纹理贴图，解决重建网格缺少真实感表面外观的问题。

核心能力

按 Waechter 2014 论文实现大规模纹理映射
输入三角网格模型与已配准图像，输出带纹理的模型
构建时自动下载编译 rayint、Eigen、MVE、mapMAP 依赖
无参数运行可查看输入格式与可选参数说明
支持 make -j 并行编译

容易踩的坑

需自行准备三角化三维模型和与之配准的图像，可用 Multi-View Environment 生成；默认构建类型为 RELWITHDEBINFO 会关闭断言，排查问题需改用 -DCMAKE_BUILD_TYPE=DEBUG 重新编译；Windows 下编译未经充分验证

内容由模型根据 README 生成，可能不完整或不准确，以官方文档为准。看原文 README

用户点评

登录用户的真实使用记录，可按版本和硬件筛选

共 0 条

还没有人点评。第一个说出真实体验的人最有价值。

README 原文

未翻译

在 GitHub 打开

MVS-Texturing

Welcome to our project that textures 3D reconstructions from images. This project focuses on 3D reconstructions generated using structure from motion and multi-view stereo techniques, however, it is not limited to this setting.

The algorithm was published in Sept. 2014 on the European Conference on Computer Vision in Let There Be Color! Large-Scale Texturing of 3D Reconstructions see publication details below.

Dependencies

The code and the build system have the following prerequisites:

cmake (>= 3.1)
git
make
gcc (>= 5.0.0) or a compatible compiler
libpng, libjpg, libtiff, libtbb

Furthermore the build system automatically downloads and compiles the following dependencies (so there is nothing you need to do here):

rayint https://github.com/nmoehrle/rayint
Eigen https://libeigen.gitlab.io
Multi-View Environment https://github.com/simonfuhrmann/mve
mapMAP https://github.com/dthuerck/mapmap_cpu

Compilation

git clone https://github.com/nmoehrle/mvs-texturing.git
cd mvs-texturing
mkdir build && cd build && cmake ..
make (or make -j for parallel compilation)

If something goes wrong during compilation you should check the output of the cmake step. CMake checks all dependencies and reports if anything is missing.

If you think that there is some problem with the build process on our side please tell us.

If you are trying to compile this under windows (which should be possible but we haven't checked it) and you feel like we should make minor fixes to support this better, you can also tell us.

Execution

As input our algorithm requires a triangulated 3D model and images that are registered against this model. One way to obtain this is to:

import images, infer camera parameters, reconstruct depth maps and fuse these depth maps into a combined 3d model using the applications from the Multi-View Environment.

By starting the application without any parameters and you will get a description of the expected file formats and optional parameters.

Troubleshooting

When you encounter errors or unexpected behavior please make sure to switch the build type to debug e.g. cmake -DCMAKE_BUILD_TYPE=DEBUG .., recompile and rerun the application. Because of the computational complexity the default build type is RELWITHDEBINFO which enables optimization but also ignores assertions. However, these assertions could give valuable insight in failure cases.

License, Patents and Citing

Our software is licensed under the BSD 3-Clause license, for more details see the LICENSE.txt file.

If you use our texturing code for research purposes, please cite our paper:

@inproceedings{Waechter2014Texturing,
  title    = {Let There Be Color! --- {L}arge-Scale Texturing of {3D} Reconstructions},
  author   = {Waechter, Michael and Moehrle, Nils and Goesele, Michael},
  booktitle= {Proceedings of the European Conference on Computer Vision},
  year     = {2014},
  publisher= {Springer},
}

Contact

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 the github repository .