【ORB_SLAM】Ubuntu20.04 配置ORB

这篇具有很好参考价值的文章主要介绍了【ORB_SLAM】Ubuntu20.04 配置ORB_SLAM3。希望对大家有所帮助。如果存在错误或未考虑完全的地方，请大家不吝赐教，您也可以点击"举报违法"按钮提交疑问。

本文主要记录基于Ubuntu20.04环境下，对普通的ORB_SLAM3和稠密版本的ORB_SLAM3进行环境的配置。

一、配置ORB_SLAM3<普通版本>

1.安装ROS开发环境

这里采用鱼香ros的一键安装，感谢小鱼的一键安装。

wget http://fishros.com/install -O fishros && . fishros

2.安装Pangolin

Pangolin：链接: https://pan.baidu.com/s/1FXYLsEK1W3xmX0m_Vqylag 提取码: jgz2

# 这里采用的是Pangolin-0.5-20.04
cd Pangolin-0.5-20.04
mkdir build
cd build
cmake ..
make -j
sudo make install

3.安装eigen

sudo apt-get install libeigen3-dev

【ORB_SLAM】Ubuntu20.04 配置ORB_SLAM3,SLAM学习,人工智能,自动驾驶,Powered by 金山文档

4.拉取orb_slam3 的源码

(基于小六的注释后的ORB_SLAM3代码）

https://github.com/electech6/ORB_SLAM3_detailed_comments

git clone https://github.com/electech6/ORB_SLAM3_detailed_comments.git

5.安装DBoW2（在Thirdparty文件中）

需要先修改opencv版本

如何查看opencv版本号?

编译（由于在工程当中，并不需要install）

cd DBoW2
mkdir build
cd build
cmake ..
make -j

6.安装g2o

cd g2o
mkdir build
cd build
cmake ..
make -j

7.开始编译ORBSLAM3（普通版）

在工程中修改CMakeLists 中opencv的版本，一定要与系统的opencv版本一致

cd ORB_SLAM3_detailed_comments
mkdir build
cd build
cmake ..
make -j

若显示以上报错，则需要安装realsense2

sudo apt-get install ros-noetic-realsense2-camera

ORB_SLAM3也可以使用一键安装

cd ORB_SLAM3_detailed_comments
sudo chmod 777 build.sh
./build.sh

8.编译ORB_SLAM3 ROS环境

同样需要先修改OpenCV的版本号为本机版本

在bashrc中添加ORBSLAM路径

sudo vim ~/.bashrc
export ROS_PACKAGE_PATH=$ROS_PACKAGE_PATH:/home/robot/orb_slam3/ORB_SLAM3_detailed_comments/Examples/ROS/ORB_SLAM3
# vim 操作方法： 输入 按a
#              结束先按Esc ，再按 :wq
# 记得source下环境
source ~/.bashrc

若出现以上错误，则需要去rosdep init一下，并且rosdep update

可以使用最上面的鱼香ros更新一下

cd ORB_SLAM3_detailed_comments/Examples/ROS/ORB_SLAM3
mkdir build
cd build
cmake ..
make -j

也可以使用一键安装

cd ORB_SLAM3_detailed_comments
sudo chmod 777 build_ros.sh
./build_ros.sh

9.测试一下

# 运行代码之前记得source一下
cd orb_slam3/ORB_SLAM3_detailed_comments/Examples/ROS/ORB_SLAM3/build
source devel/setup.bash

# 终端1
roscore
# 终端2
rosrun ORB_SLAM3 Mono_Inertial /home/robot/orb_slam3/ORB_SLAM3_detailed_comments/Vocabulary/ORBvoc.txt /home/robot/orb_slam3/ORB_SLAM3_detailed_comments/Examples/Monocular-Inertial/EuRoC.yaml

测试数据集：

链接: https://pan.baidu.com/s/1WwTSvQqJ8QSpPPs0G-dXNg 提取码: nnej

rosbag play V1_01_easy.bag

这样,普通版本的ORB_SLAM3环境搭建成功！

将ORB-SLAM3编写成launch文件

<?xml version="1.0"?>
<launch>
    <node pkg="ORB_SLAM3" type="RGBD" name="RGBD" args="/home/qsx/orb_slam/ORB_SLAM3_detailed_comments/Vocabulary/ORBvoc.txt /home/qsx/orb_slam/ORB_SLAM3_detailed_comments/Examples/RGB-D/RealSense_L515.yaml" />

    <arg name="node_start_delay" default="5.0" />  
    <node pkg="rosbag" type="play" name="rosbag_play" args="/home/qsx/F1_dataset/1/record_12_6_F1_3.bag" launch-prefix="bash -c 'sleep $(arg node_start_delay); $0 $@' " />
</launch>

二、编译ORB_SLAM3(稠密版本)

1.下载源码：

git clone -b dense_map_new https://github.com/electech6/ORB_SLAM3_detailed_comments.git

同样需要先安装Thirdparty中的DBoW2和g2o，方法同上。

2.修改opencv版本

若出现以上错误，则需要安装pcl-1.12版本

3.解决pcl多版本共存问题

pcl各个版本下载地址：https://github.com/PointCloudLibrary/pcl/releases

下载pcl-1.12.1：https://github.com/PointCloudLibrary/pcl/archive/refs/tags/pcl-1.12.1.zip

(个人认为：由于我们一般编译安装的文件都放再/usr/local路径，而若是我们自行定义自己的文件夹，这样系统就找不到，也就不会出现多版本共存冲突的问题啦！<俗话说，打不过我躲还行吧！>但需要在CMakeList文件中去寻找我们的文件路径)

①将下载的1.12.1源码放到pcl-1.12文件夹中，再新建一个pcl-1.12_installfile文件用来存放安装后的文件

②开始编译pcl

cd pcl-1.12
mkdir build
cd build
cmake -DCMAKE_BUILD_TYPE=None -DCMAKE_INSTALL_PREFIX=/home/robot/Documents/pcl-1.12_installfile ..
make
make install 
make clean

③修改CmakeList文件

修改成如下：

find_package(PCL 1.12 REQUIRED 
PATHS /home/robot/Documents/pcl-1.12_installfile
)

4.修改c++版本

若出现以上错误，则需要把C++11改为C++14

把C++11改为C++14

5.C++问题

若出现error: ‘std::chrono::monotonic_clock’ has not been declared

则全部替换成steady_clock

这样就可以编译通过啦！

6.编译稠密版本的ORB_SLAM3 ROS

①添加bashrc环境

export ROS_PACKAGE_PATH=$ROS_PACKAGE_PATH:/home/robot/orb_slam3/dense_ORB_SLAM3_detailed_comments/Examples/ROS/ORB_SLAM3

# 保存完记得 
source ~/.bashrc

②修改CMakeList文件

③使用一键安装

sudo chmod 777 build_ros.sh
./build_ros.sh

④pcl问题

若出现pcl/point_types.h:No such file or directory

则需要修改CMakeList文件

⑤C++版本问题

若出现上述问题，则需要将C++11改为C++14

这样就ROS的稠密版本也编译完成啦！

7.测试一下

①编写launch文件

<?xml version="1.0"?>
<launch>
    <!-- <include file="$(find ORB_SLAM3)/launch/rs_camera.launch"/> -->


    <node pkg="ORB_SLAM3" type="RGBD" name="RGBD" args="/home/robot/orb_slam3/dense_map_ORB_SLAM3_detailed_comments/Vocabulary/ORBvoc.txt /home/robot/orb_slam3/dense_map_ORB_SLAM3_detailed_comments/Examples/ROS/ORB_SLAM3/config/RealSense_L515.yaml" output="screen"/>

    <arg name="node_start_delay" default="5.0" />  

    <node pkg="rosbag" type="play" name="rosbag_play" args="/media/robot/Q小鑫/数据集/R2DIO_data/F3_scene3_data/F3_scene3_dataset/F3_scene3_4.bag" launch-prefix="bash -c 'sleep $(arg node_start_delay); $0 $@' " />

    
</launch>

②修改相机Yaml文件（采用的是RealSense_L515相机）

%YAML:1.0

#--------------------------------------------------------------------------------------------
# Camera Parameters. Adjust them!
#--------------------------------------------------------------------------------------------
# File.version: "1.0"

Camera.type: "PinHole"

# Right Camera calibration and distortion parameters (OpenCV)
Camera.fx: 603.086
Camera.fy: 603.23
Camera.cx: 329.578
Camera.cy: 232.509


# distortion parameters
Camera.k1: 0.15858
Camera.k2: -0.492513
Camera.k3: 0.4519
Camera.p1: -0.00218088
Camera.p2: 0.00021294 

# Camera resolution
Camera.width: 640
Camera.height: 480

# Camera frames per second 
Camera.fps: 30

# IR projector baseline times fx (aprox.)
Camera.bf: 30.33325

# Color order of the images (0: BGR, 1: RGB. It is ignored if images are grayscale)
Camera.RGB: 1

# Close/Far threshold. Baseline times.
ThDepth: 40.0
#Stereo.b: 0.0745
DepthMapFactor: 1000.0
# DepthMapFactor: 1.0
# Depth map values factor
# RGBD.DepthMapFactor: 1000.0

#--------------------------------------------------------------------------------------------
# ORB Parameters
#--------------------------------------------------------------------------------------------
# ORB Extractor: Number of features per image
ORBextractor.nFeatures: 1000

# ORB Extractor: Scale factor between levels in the scale pyramid     
ORBextractor.scaleFactor: 1.2

# ORB Extractor: Number of levels in the scale pyramid    
ORBextractor.nLevels: 8

# ORB Extractor: Fast threshold
# Image is divided in a grid. At each cell FAST are extracted imposing a minimum response.
# Firstly we impose iniThFAST. If no corners are detected we impose a lower value minThFAST
# You can lower these values if your images have low contrast            
ORBextractor.iniThFAST: 20
ORBextractor.minThFAST: 7

#--------------------------------------------------------------------------------------------
# Viewer Parameters
#--------------------------------------------------------------------------------------------
Viewer.KeyFrameSize: 0.05
Viewer.KeyFrameLineWidth: 1.0
Viewer.GraphLineWidth: 0.9
Viewer.PointSize: 2.0
Viewer.CameraSize: 0.08
Viewer.CameraLineWidth: 3.0
Viewer.ViewpointX: 0.0
Viewer.ViewpointY: -0.7
Viewer.ViewpointZ: -1.8
Viewer.ViewpointF: 500.0

PointCloudMapping.Resolution: 0.05
meank: 50
thresh: 1.0