高版本Ubuntu(22.04/20.04)安装ORB-SLAM2并试运行

本文最后更新于：2022年9月12日 晚上

ubuntu 20.04 篇

  1. install Pangolin, not follow the github guild, follow me
    1. sudo apt install libgl1-mesa-dev
    2. sudo apt install libglew-dev
    3. sudo apt install cmake
    4. download Pangolin and cd Pangolin
    5. mkdir build && cd build
    6. sudo apt install libeigen3-dev
    7. cmake ..
    8. cmake –build .
    9. sudo make install
  2. install opencv-3.4.6
    1. sudo apt install build-essential
    2. sudo apt install libgtk2.0-dev
    3. sudo apt install libavcodec-dev
    4. sudo apt install libavformat-dev
    5. sudo apt install libjpeg.dev (maybe libjpef-dev also working)
    6. download opencv-3.4.6 source
    7. cd opencv
    8. mkdir build && cd build
    9. cmake ..
    10. make -j8 (so many warnings, but no error)
    11. sudo make install
  3. build ORB-SLAM2 (https://github.com/raulmur/ORB_SLAM2)
    1. git clone https://github.com/raulmur/ORB_SLAM2.git ORB_SLAM2
    2. sudo apt install libcanberra-gtk-module
    3. cd ORB_SLAM2/
    4. chmod +x build.sh
    5. sed -i ‘s/++11/++14/g’ CMakeLists.txt
    6. add add_definitions(-w) to CMakeLists.txt
    7. change find_package(Eigen3 3.1.0 REQUIRED) to find_package(Eigen3 3.1.0 REQUIRED NO_MODULE)
    8. add #include <unistd.h> to LoopClosing.h
    9. edit LoopClosing.h
      1
      typedef map<KeyFrame*,g2o::Sim3,std::less<KeyFrame*>,Eigen::aligned_allocator<std::pair<const KeyFrame*, g2o::Sim3> > > KeyFrameAndPose;
      change to
      1
      typedef map<KeyFrame*,g2o::Sim3,std::less<KeyFrame*>,Eigen::aligned_allocator<std::pair<KeyFrame *const, g2o::Sim3> > > KeyFrameAndPose;
    10. ./build.sh
  4. run ORB-SLAM2
    1. sudo ldconfig
    2. ./Examples/Monocular/mono_tum Vocabulary/ORBvoc.txt Examples/Monocular/TUM1.yaml /home/conan/slam2/datasets/rgbd_dataset_freiburg1_desk
    3. install anaconda
    4. create py27 env
    5. activate py27 and pip install numpy
    6. download associate.py and move to datasets, cd datasets path
    7. python associate.py rgb.txt depth.txt > associations.txt
    8. ./Examples/RGB-D/rgbd_tum Vocabulary/ORBvoc.txt Examples/RGB-D/TUM1.yaml /home/conan/slam2/datasets/rgbd_dataset_freiburg1_desk /home/conan/slam2/datasets/rgbd_dataset_freiburg1_desk/associations.txt

Ubuntu 22.04 篇

  1. install Pangolin, not follow the github guild, follow me
    1. sudo apt install libgl1-mesa-dev
    2. sudo apt install libglew-dev
    3. sudo apt install cmake
    4. download Pangolin and cd Pangolin
    5. mkdir build && cd build
    6. manual install eigen3(3.3.7), do not use the apt to install eigen3 (like: apt install libeigen3-dev, in 22.04, the version of eigen is 3.4.0, does not fit for plangolin)
    7. cmake ..
    8. cmake –build .
    9. sudo make install
  2. install opencv-3.4.6
    1. sudo apt install build-essential
    2. sudo apt install libgtk2.0-dev
    3. sudo apt install libavcodec-dev
    4. sudo apt install libavformat-dev
    5. sudo apt install libjpeg-dev
    6. download opencv-3.4.6 source
    7. cd opencv
    8. mkdir build && cd build
    9. cmake ..
    10. make sure change the gcc version to 9, ubuntu 22.04 default gcc version is 11
    11. make -j8 (so many warnings, but no error)
    12. sudo make install
  3. build ORB-SLAM2 (https://github.com/raulmur/ORB_SLAM2)
    1. git clone https://github.com/raulmur/ORB_SLAM2.git ORB_SLAM2
    2. sudo apt install libcanberra-gtk-module
    3. cd ORB_SLAM2/
    4. chmod +x build.sh
    5. sed -i ‘s/++11/++14/g’ CMakeLists.txt
    6. add add_definitions(-w) to CMakeLists.txt
    7. change find_package(Eigen3 3.1.0 REQUIRED) to find_package(Eigen3 3.1.0 REQUIRED NO_MODULE)
    8. add #include <unistd.h> to LoopClosing.h
    9. edit LoopClosing.h
      1
      typedef map<KeyFrame*,g2o::Sim3,std::less<KeyFrame*>,Eigen::aligned_allocator<std::pair<const KeyFrame*, g2o::Sim3> > > KeyFrameAndPose;
      change to
      1
      typedef map<KeyFrame*,g2o::Sim3,std::less<KeyFrame*>,Eigen::aligned_allocator<std::pair<KeyFrame *const, g2o::Sim3> > > KeyFrameAndPose;
    10. ./build.sh
  4. run ORB-SLAM2
    1. sudo ldconfig
    2. ./Examples/Monocular/mono_tum Vocabulary/ORBvoc.txt Examples/Monocular/TUM1.yaml /home/conan/slam2/datasets/rgbd_dataset_freiburg1_desk
    3. install anaconda
    4. create py27 env
    5. activate py27 and pip install numpy
    6. download associate.py and move to datasets, cd datasets path
    7. python associate.py rgb.txt depth.txt > associations.txt
    8. ./Examples/RGB-D/rgbd_tum Vocabulary/ORBvoc.txt Examples/RGB-D/TUM1.yaml /home/conan/slam2/datasets/rgbd_dataset_freiburg1_desk /home/conan/slam2/datasets/rgbd_dataset_freiburg1_desk/associations.txt
视觉SLAM
#Ubuntu #ORB-SLAM2
高版本Ubuntu(22.04/20.04)安装ORB-SLAM2并试运行
https://cnwjy.site/2022/09/01/高版本Ubuntu-22-04-20-04-安装ORB-SLAM2并试运行/
作者
Conan-jine
发布于
2022年9月1日
许可协议