Link to original ORB-SLAM3's README.md
Link to thien94's README.md Had already implemented many updates
Link to Mauhing's fork Other fixes, some incorporated here.
disclaimer: I started my journey in the world of Ubuntu and ROS May 2022 (i.e. a noob). These instructions are written mainly for myself as I learn. If in doubt (as you should be) follow this instead: shanpenghui
By following these instructions you will be able to run ORB SLAM 3 with ROS Noetic on a fresh install of Ubuntu 20.04.5 LTS, publish your camera on a ros node and run SLAM on the included datasets.
In a future update this will also cover running SLAM on your own dataset, including calibration of your camera.
Make sure your system is up to date. Open a terminal [ctrl + alt + t] and run
sudo apt update && sudo apt upgrade
#restart if necessary
Some of the packages we are building require a certain amount of memory. If you are running Ubuntu in a virtual environment you should allocate 12Gb of memory to Ubuntu or install Ubuntu as a dual boot to give it all your RAM. If you only have say 8 Gb available I encourage you to follow these steps to increase your swap disk (think of it as fake RAM using HDD space) ask Ubuntu - increase Swap. I've pasted the commands here:
swapon --show
Disable and delete it:
sudo swapoff /swapfile
sudo rm /swapfile
Don't create a swap bigger than the amount of RAM available as this might cause issues. Do simple math: set swap to say 8Gb (8 * 1024 = 8192) or 16Gb (16 * 1024 = 16384).
#The following command will allocate 8Gb for swap:
sudo dd if=/dev/zero of=/swapfile bs=1M count=8192
grant root permissions and format as swap:
sudo chmod 600 /swapfile
sudo mkswap /swapfile
Be patient, my machine froze for a while during this step.
sudo swapon /swapfile
and reboot for it to take effect. To verify repeat the top command
reboot
Run the following commands in terminal
sudo add-apt-repository "deb http://security.ubuntu.com/ubuntu xenial-security main"
sudo apt update
sudo apt-get install -y build-essential cmake git libgtk2.0-dev pkg-config libavcodec-dev libavformat-dev libswscale-dev python-dev python-numpy libtbb2 libtbb-dev libjpeg-dev libpng-dev libtiff-dev libdc1394-22-dev libjasper-dev libglew-dev libboost-all-dev libssl-dev
sudo apt install libeigen3-dev curl
mkdir ~/Dev
cd ~/Dev
git clone https://github.com/stevenlovegrove/Pangolin.git
cd Pangolin
mkdir build && cd build
The source code is now on your machine and the following command will create building instructions based on dependencies listed in CMakeLists.txt (~/Dev/Pangolin/CMakeLists.txt)
cmake ..
The -j4 specifies how many jobs you would like to process simultaniously. If the build fails, try again with "-j1" before turning to google.
make -j4
If successful finish the installation by running:
sudo make install
Check the OpenCV version on your computer (required at leat 3.0 as stated in the original README.md):
python3 -c "import cv2; print(cv2.__version__)"
On a freshly installed Ubuntu 20.04.4 LTS with desktop image, OpenCV 4.2.0 is included so we can skip to the next step. If a newer version is required (>= 3.0), follow the instrucions:
- General installation instruction.
- If you want CUDA to be included: How to install OpenCV 4.5.2 with CUDA 11.2 and CUDNN 8.2 in Ubuntu 20.04
For example, the main commands for OpenCV 4.4.0 (without CUDA):
cd ~/Dev
git clone https://github.com/opencv/opencv
git -C opencv checkout 4.4.0
gedit ~/Dev/opencv/modules/videoio/src/cap_ffmpeg_impl.hpp
A script just opened, now add the following lines in the top:
#define AV_CODEC_FLAG_GLOBAL_HEADER (1 << 22)
#define CODEC_FLAG_GLOBAL_HEADER AV_CODEC_FLAG_GLOBAL_HEADER
#define AVFMT_RAWPICTURE 0x0020
Then continue as usual:
mkdir -p ~/Dev/opencv/build
cd ~/Dev/opencv/build
cmake ..
make -j4
sudo make install
Clone the repo like this:
cd ~/Dev
git clone https://github.com/discoimp/ORB_SLAM3.git ORB_SLAM3
Run the shell file containing several build commands.
cd ORB_SLAM3
chmod +x build.sh
./build.sh
cd ~
mkdir -p Datasets/EuRoC
cd Datasets/EuRoC/
wget -c http://robotics.ethz.ch/~asl-datasets/ijrr_euroc_mav_dataset/machine_hall/MH_01_easy/MH_01_easy.zip
mkdir MH01
unzip MH_01_easy.zip -d MH01/
If you later intend to use ROS with the examples, just run this now to point it in the correct direction:
echo "export ROS_PACKAGE_PATH=${ROS_PACKAGE_PATH}:~Dev/ORB_SLAM3/Examples_old/ROS" >> ~/.bashrc
if the following examples fail with something related to Pangolin. I got "./Examples/Monocular/mono_euroc: error while loading shared libraries: libpango_windowing.so: cannot open shared object file: No such file or directory" then go to the bottom of this README to find one possible solution:
cd ~/Dev/ORB_SLAM3
./Examples/Monocular-Inertial/mono_inertial_euroc ./Vocabulary/ORBvoc.txt ./Examples/Monocular-Inertial/EuRoC.yaml ~/Datasets/EuRoC/MH01 ./Examples/Monocular-Inertial/EuRoC_TimeStamps/MH01.txt dataset-MH01_monoimu
For now I'm tweaking this:rpg_dvs_ros
Then learn how to calibrate your own camera: First publish a camera node in ros: https://automaticaddison.com/working-with-ros-and-opencv-in-ros-noetic/ Then Calibrate it: http://wiki.ros.org/camera_calibration/Tutorials/MonocularCalibration
This is a bit more advanced, but not impossible.
I ran into trouble running the dataset and received the same error described by someone else here: issue 399 First check if the path to your missing component is referenced in this: “/etc/ld.so.conf" (follow and open the paths listed)
gedit /etc/ld.so.conf
then run this to flush the cache (I guess)
sudo ldconfig
if the library still is not working we can do this unsactioned move: homebrew install instruction I've pasted the [relevant] commands here (given you have completed the steps above): WARNING: If you google "use several package managers Linux" or simular, it comes out as a bad idea if you don't know what you're doing. So this seems like a risky move, but it worked for me (so far...).
sudo apt update
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
The above command may take a while Then tell the system about it's new package manager super skill:
eval "$(/home/linuxbrew/.linuxbrew/bin/brew shellenv)"
Check if all is okay
brew doctor
then:
brew install gcc
package manger should now be installed
Now we can check dependencies of Pangolin using brew. If I understand this correctly this may override some of the installations you already have going, and migth lead to complex epic failures.
cd ~/Dev/Pangolin/
./scripts/install_prerequisites.sh -m brew all
If it fails, try running it again. I had to run it twice.
cmake -B build
cmake --build build
thien94 ros addon have not tested yet.
Install guide Davis ++ drivers -includes setting up a catkin build workspace