随笔- 1525 文章- 2 评论- 66 阅读- 127万

https://github.com/HKUST-Aerial-Robotics/VINS-Fusion

https://blog.csdn.net/haner27/article/details/104004745

0环境

opencv3.4.9

ubuntu20

1.ROS安装

我自己的是ubuntu20

安装ros后，安装需要补充的ROS 功能包：

1	`sudo apt-get` `install ros-melodic-cv-bridge ros-melodic-tf ros-melodic-message-filters ros-melodic-image-transport`

1.2 修改ros cv bridge

https://blog.csdn.net/zhaodeming000/article/details/119859665

ros自带的 bridge 是opencv4的，自己的是opencv3.4.9,需要单独重新编译下

究其原因，是因为在安装ROS时，cv_bridge默认连接了4.0+版本的opencv进行编译。只修改cvbridgeConfig.cmake中opencv的路径，编译可能没问题，但在代码执行时如果引用到libcv_bridge.so，还是会链接到opencv原来的版本。

1.卸载原来版本的cv_bridge

sudo apt-get remove ros-noetic-cv-bridge

2.下载新版本cv_bridge

浏览器打开
https://github.com/ros-perception/vision_opencv
选择noetic对应的cv_bridge，
或者直接，
git clone https://github.com/ros-perception/vision_opencv/tree/melodic

3.修改cv_bridge中的cmakelists.txt

find_package(OpenCV 3.4.10 REQUIRED) //改成你安装的opencv的版本

4.cmake单独编译cv_bridge

在cv_bridge文件夹下，

mkdir build
 
cd build
 
cmake ..
 
sudo make
 
sudo make install

5.在开源代码需要的cmaklist.txt中添加cv_bridge的cmake路径

set(cv_bridge_DIR /usr/local/share/cv_bridge/cmake) //在find_package前面
比如，在vinsfusion开源SLAM代码中，只需要在loop_fusion和vins_estimator两个子模块的cmakelist.txt开头添加set，即可完美编译通过。

2.ceres solver安装

2.安装ceres1.14.0

接着从以下链接安装ceres1.14.0：

https://codeload.github.com/ceres-solver/ceres-solver/tar.gz/refs/tags/1.14.0

http://ceres-solver.org/installation.html

(之前openvslam 安装了cere版本直接编译报错，大概率不适配，这里安装最新的，以后如果要开启openvslam 记得还原对应ceres版本重装 sudo make uninstall)

官方给的源是国外源，下载不了，这里是国内的下载链接
https://github.com/ceres-solver/ceres-solver.git

# CMake
sudo apt-get install cmake
# google-glog + gflags
sudo apt-get install libgoogle-glog-dev libgflags-dev
# Use ATLAS for BLAS & LAPACK
sudo apt-get install libatlas-base-dev
# Eigen3
sudo apt-get install libeigen3-dev
# SuiteSparse (optional)
sudo apt-get install libsuitesparse-dev

#14.04
sudo apt-get install liblapack-dev libsuitesparse-dev libcxsparse3.1.2 libgflags-dev libgoogle-glog-dev libeigen3-dev libgtest-dev
#18.04 20.04
sudo apt-get install liblapack-dev libsuitesparse-dev libcxsparse3 libgflags-dev libgoogle-glog-dev libeigen3-dev libgtest-dev
 
git clone https://github.com/ceres-solver/ceres-solver.git
cd ceres-solver/
mkdir build
cd build
cmake ..
make
sudo make install

如果报错cuda错误，大概率是找不到cuda的位置，手动添加cmakelist.txt

1	`set(CMAKE_CUDA_COMPILER /usr/local/cuda/bin/nvcc)`

　　记得安装依赖项，记得最后的make install

3.VINS_Fusion安装

创建ros空间

　　安装VINS_Fusion

cd ~/catkin_ws/src
git clone https://github.com/HKUST-Aerial-Robotics/VINS-Fusion.git
cd ../
catkin_make
source ~/catkin_ws/devel/setup.bash

注：直接git获取VINS_Fusion，由于网络原因获取不成功，换了个电脑在github网站下载后，放到/catkin_ws/src文件夹中即可。

cd ~/catkin_ws
catkin_make
source ~/catkin_ws/devel/setup.bash

opencv报错解决

1 添加头文件

cv::CV_GRAY2RGB 报错替换为

COLOR_GRAY2RGB

对应文件添加

1	`#include<opencv2/imgproc/types_c.h>`

2 一些老的参数换成新的参数

CV_FONT_HERSHEY_SIMPLEX报错替换
cv::FONT_HERSHEY_SIMPLEX
 
CV_LOAD_IMAGE_GRAYSCALE报错替换
cv::LOAD_IMAGE_GRAYSCALE
 
CV_LOAD_IMAGE_GRAYSCALE报错替换
cv::ImreadModes::IMREAD_GRAYSCALE

3跑通实例

相机参数

设置输出参数

数据位置

#!/bin/sh
  
#延迟5秒执行
sleep 1
echo "运行开始"
#可视化
cd /home/dongdong/v2_Project/v3_SLAM/v5_VinsFusion/catkin
source devel/setup.bash
roslaunch vins vins_rviz.launch
 
# 实时双目视觉vo
cd /home/dongdong/v2_Project/v3_SLAM/v5_VinsFusion/catkin
source devel/setup.bash
rosrun vins kitti_gps_test /home/dongdong/v2_Project/v3_SLAM/v5_VinsFusion/catkin/src/VINS-Fusion/config/kitti_raw/kitti_10_03_config.yaml /media/dongdong/Elements/数据集/公共数据集/vinskkt数据/2011_10_03_drive_0027_sync 
 
# gps全局优化
cd /home/dongdong/v2_Project/v3_SLAM/v5_VinsFusion/catkin
source devel/setup.bash
rosrun global_fusion global_fusion_node

自动脚本

给与脚本权限

1	`sudo chmod -R 777 run.sh`

　　脚本内容

#!/bin/sh
  
#延迟5秒执行
sleep 1
echo "运行开始"
 
gnome-terminal -t "1-vins_rviz" -x bash -c "\
cd /home/dongdong/v2_Project/v3_SLAM/v5_VinsFusion/catkin ;\
source devel/setup.bash ;\
roslaunch vins vins_rviz.launch ;\
exit;exec bash;"
 
sleep 3
 
echo "2 kitti_gps_test"
 
 
gnome-terminal -t "2-kitti_gps_test" -x bash -c "\
cd /home/dongdong/v2_Project/v3_SLAM/v5_VinsFusion/catkin ;\
source devel/setup.bash ;\
rosrun vins kitti_gps_test /home/dongdong/v2_Project/v3_SLAM/v5_VinsFusion/catkin/src/VINS-Fusion/config/kitti_raw/kitti_10_03_config.yaml /media/dongdong/Elements/数据集/公共数据集/vinskkt数据/2011_10_03_drive_0027_sync ;\ 
exit;exec bash;"
sleep 1
 
echo "3 global_fusion"
 
gnome-terminal -t "3-global_fusion" -x bash -c "\
cd /home/dongdong/v2_Project/v3_SLAM/v5_VinsFusion/catkin ;\
source devel/setup.bash ;\
rosrun global_fusion global_fusion_node ;\
exit;exec bash;"
sleep 1

4.KITTI数据集下载

http://www.cvlibs.net/datasets/kitti/raw_data.php
国外源打不开，感谢这个博主提供的国内源下载：https://blog.csdn.net/weixin_43599336/article/details/86533572
（刚开始可以下载，后来中断了，网页打不开，等一会再下就可以）

泡泡链接（百度网盘）：https://www.sohu.com/a/219232053_715754

这两个链接都没能成功下载，最后找到了下面的链接：
https://www.dalipan.com/detail/40c8f99ba2120a493a148f8332094668

终于成功下载了。。

5.跑通KITTI数据集
1.KITTI Odometry (Stereo)：使用sequences 00数据集

roslaunch vins vins_rviz.launch
(optional) rosrun loop_fusion loop_fusion_node ~/catkin_ws/src/VINS-Fusion/config/kitti_odom/kitti_config00-02.yaml
rosrun vins kitti_odom_test ~/catkin_ws/src/VINS-Fusion/config/kitti_odom/kitti_config00-02.yaml YOUR_DATASET_FOLDER/sequences/00/ 