ORB-SLAM3 Installation Guide

This guide provides step-by-step instructions for installing ORB-SLAM3 along with its dependencies. Ensure that you meet the prerequisites before starting the installation process.

Prerequisites

• OpenCV version greater than 4.4.0 (specifically, 4.5.0 is recommended)

Install OpenCV

1. Open a terminal and create a directory for development :

   cd ~
   mkdir Dev && cd Dev
   

2. Clone the OpenCV repository:

   git clone https://github.com/opencv/opencv.git
   cd opencv
   git checkout 4.5.0

3. Modify the cap_ffmpeg_impl.hpp header file:

   gedit ./modules/videoio/src/cap_ffmpeg_impl.hpp

   Add the following lines at the top of the file:

   #define AV_CODEC_FLAG_GLOBAL_HEADER (1 << 22)
   #define CODEC_FLAG_GLOBAL_HEADER AV_CODEC_FLAG_GLOBAL_HEADER
   #define AVFMT_RAWPICTURE 0x0020

   Save and close the file.

4. Create a build directory and compile OpenCV:

   mkdir build
   cd build
   cmake -D CMAKE_BUILD_TYPE=Release -D WITH_CUDA=OFF -D CMAKE_INSTALL_PREFIX=/usr/local ..
   make -j 3
   sudo make install

Get Pangolin

1. Navigate to your preferred code directory:

   cd ~/your_fav_code_directory

2. Clone the Pangolin repository:

   git clone --recursive https://github.com/stevenlovegrove/Pangolin.git
   cd Pangolin

3. Install dependencies:

   ./scripts/install_prerequisites.sh recommended

4. Configure and build Pangolin:

   cmake -B build
   cmake --build build

Install ORB-SLAM3

1. Clone the ORB-SLAM3 repository:

   git clone https://github.com/ans-qureshi/ORB_SLAM3
   cd ORB_SLAM3

2. Build ORB-SLAM3:

   ./build.sh

Install ORB-SLAM3 with ROS

If you want to use ORB-SLAM3 with ROS, run:

./build_ros.sh

Conclusion

You have successfully installed ORB-SLAM3 along with its dependencies. Make sure to follow any additional setup instructions specific to your project requirements. Enjoy using ORB-SLAM3! If you use ORB-SLAM3 in an academic work, please cite:

    @article{ORBSLAM3_TRO,
      title={{ORB-SLAM3}: An Accurate Open-Source Library for Visual, Visual-Inertial 
               and Multi-Map {SLAM}},
      author={Campos, Carlos AND Elvira, Richard AND G\´omez, Juan J. AND Montiel, 
              Jos\'e M. M. AND Tard\'os, Juan D.},
      journal={IEEE Transactions on Robotics}, 
      volume={37},
      number={6},
      pages={1874-1890},
      year={2021}
     }