ORB-SLAM3 Installation Guide for NVIDIA Jetson AGX Xavier

This guide is designed to help you install ORB-SLAM3 on a NVIDIA Jetson AGX Xavier board running Jetpack (which is based on Ubuntu 20.04). The Xavier AGX platform has an ARM architecture, so make sure to use compatible libraries and packages.

Prerequisites

Before you begin, ensure your Xavier board is running the latest version of Jetpack. You can download and install it from the NVIDIA official website.

Step 1: Install Dependencies

First, update your package list and install the necessary dependencies:

sudo apt update && sudo apt upgrade -y
sudo apt-get install build-essential cmake git libgtk-3-dev pkg-config libavcodec-dev libavformat-dev libswscale-dev
sudo apt-get install python3-dev python3-numpy libtbb2 libtbb-dev libjpeg-dev libpng-dev libtiff-dev libdc1394-22-dev
sudo apt-get install libeigen3-dev libglew-dev libboost-all-dev libssl-dev

Step 2: Install OpenCV

ORB-SLAM3 requires OpenCV. Let's build OpenCV from source:

# Create a directory for the OpenCV build
cd ~
mkdir -p Dev && cd Dev
mkdir opencv_build && cd opencv_build

# Clone the OpenCV's and OpenCV contrib repositories
git clone https://github.com/opencv/opencv.git
git clone https://github.com/opencv/opencv_contrib.git

# Checkout a stable version, for example, 4.5.0
cd opencv && git checkout 4.5.0
cd ../opencv_contrib && git checkout 4.5.0

# Prepare the build
cd ../opencv
mkdir build && cd build

# Configure the OpenCV build with CMake
cmake -D CMAKE_BUILD_TYPE=RELEASE \
      -D CMAKE_INSTALL_PREFIX=/usr/local \
      -D OPENCV_EXTRA_MODULES_PATH=../../opencv_contrib/modules \
      -D EIGEN_INCLUDE_PATH=/usr/include/eigen3 \
      -D WITH_CUDA=ON \
      -D CUDA_ARCH_BIN=7.2 \
      -D CUDA_ARCH_PTX="" \
      -D WITH_CUDNN=ON \
      -D WITH_CUBLAS=1 \
      -D ENABLE_FAST_MATH=1 \
      -D CUDA_FAST_MATH=1 \
      -D ENABLE_NEON=ON \
      -D WITH_QT=OFF \
      -D WITH_OPENCL=OFF \
      -D BUILD_TESTS=OFF \
      -D BUILD_PERF_TESTS=OFF \
      -D BUILD_EXAMPLES=OFF ..

# Build OpenCV
make -j$(nproc)
sudo make install

Step 3: Install Pangolin

cd ~/Dev

# Clone the Pangolin repository
git clone https://github.com/stevenlovegrove/Pangolin.git

# Enter the Pangolin directory
cd Pangolin

# Checkout the specific commit for Pangolin if required
git checkout 86eb4975fc4fc8b5d92148c2e370045ae9bf9f5d

# Create a build directory for Pangolin and enter it
mkdir build && cd build

# Configure the Pangolin build with CMake for release
cmake .. -DCMAKE_BUILD_TYPE=Release

# Compile Pangolin using the available cores on Xavier AGX (use 'nproc' to determine and use all available cores)
make -j$(nproc)

# Install Pangolin on your system
sudo make install

Step 4: Clone and build ORB-SLAM3

cd ~/Dev
git clone https://github.com/UZ-SLAMLab/ORB_SLAM3.git ORB_SLAM3
cd ORB_SLAM3
git submodule update --init --recursive
chmod +x build.sh
./build.sh

Step 5: Running ORB-SLAM3 Live with RealSense D435 Camera

To run ORB-SLAM3 with a RealSense D435 camera, follow these steps:

Plug in the RealSense D435 camera to the AGX board.
Navigate to the ORB-SLAM3 directory:
```
cd ~/ORB_SLAM3
```

Execute the following command:

./Examples/Stereo/stereo_realsense_D435i Vocabulary/ORBvoc.txt ./Examples/Stereo/RealSense_D435i.yaml

Handling Segmentation Fault Error

If you encounter a segmentation fault error on Step 5, such as:

SLAM settings:
    -Camera 1 parameters (Pinhole): [ 382.613 382.613 320.183 236.455 ]
Segmentation fault (core dumped)

You can resolve it by commenting out a specific line in the ORB_SLAM3/src/System.cc file. Open the file and comment out the following line:

// cout << (*settings_) << endl;

Then recompile ORB-SLAM3:

cd ~/Dev/ORB_SLAM3
./build.sh

and go back to Step 5.

NirajPudasaini/ORB_SLAM3