Colmap的安装和使用

Colmap

• 网站: https://colmap.github.io/
• GitHub: https://github.com/colmap/colmap

安装

Windows

从 GitHub发布页 GitHub Releases 下载预编译的二进制, 区分带CUDA和不带CUDA的版本.

Ubuntu

在 Ubuntu 22.04 下可以通过apt install colmap安装, 但是这样安装的是不带CUDA支持的版本

支持CUDA的版本需要通过编译安装, 可以参考的安装说明

• https://github.com/colmap/colmap/issues/2366
• https://github.com/dberga/nerfstudio/blob/main/INSTALL.md

注意不要被 Conda 的环境影响

Make sure you configure and compile from a consistent dev environment. It seems you are using anaconda. You probably want to start from a clean build folder and make sure you are not in a virtual anaconda environment.

编译步骤 (参考 https://github.com/dberga/nerfstudio/blob/main/INSTALL.md )

1. 安装依赖

sudo apt-get install \
    git \
    cmake \
    ninja-build \
    build-essential \
    libboost-program-options-dev \
    libboost-filesystem-dev \
    libboost-graph-dev \
    libboost-system-dev \
    libeigen3-dev \
    libflann-dev \
    libfreeimage-dev \
    libmetis-dev \
    libgoogle-glog-dev \
    libgtest-dev \
    libsqlite3-dev \
    libglew-dev \
    qtbase5-dev \
    libqt5opengl5-dev \
    libcgal-dev \
    libceres-dev

导出仓库

git clone https://github.com/colmap/colmap.git

编译, 注意CMAKE_CUDA_ARCHITECTURES代表的是显卡硬件架构编号, rtx2080ti 对应75, rtx4060ti 对应89

cd colmap
mkdir build
cd build
cmake .. -GNinja -DCMAKE_CUDA_ARCHITECTURES=75
sudo chown -R $(whoami) .
ninja -j1 
sudo ninja install

使用

• https://www.youtube.com/watch?v=pvvnPib4lFA
• https://www.youtube.com/watch?v=mUDzWCuopBo
• https://www.youtube.com/watch?v=Zm1mkOi9_1c&pp=ygUGY29sbWFw
• https://zxl19.github.io/colmap-note/