SLAM各算法运行方法与过程

本文介绍本人在实践过程中遇到的各个间接法的运行和配置过程。

主要包括以下五种算法:

1. VINS_mono/fusion
2. OKVIS
3. ROVIO
4. VI_ORB-SLAM
5. MSCKF

VINS_mono

---

OKVIS

Run OKVIS

可能需要加入std::ftream
Opencv需要3.4.2或以下版本，需要opencvv模块
Opencv3.4.7没有这个模块

---

ROVIO

可参考：[该博客]https://www.cnblogs.com/Jessica-jie/p/6607719.html

catkin build rovio --cmake-args -DCMAKE_BUILD_TYPE=Release -DMAKE_SCENE=ON

运行ROVIO

$ source devel/setup.bash 
$ roslaunch rovio rovio_node.launch

修改代码以输出路径

在发送IMU数据下边添加输出到文件的代码

//把数据写到文档里
 std::ofstream vio_result_file("/home/guoben/Documents/output/vio_result.csv", ios::app);
 vio_result_file << ros::Time(mpFilter_->safe_.t_) << " " 
 << imuOutput_.WrWB()(0) << " " 
 << imuOutput_.WrWB()(1) << " " 
 << imuOutput_.WrWB()(2)  << " " 
 << imuOutput_.qBW().x()  << " " 
 << imuOutput_.qBW().y()<< " " 
 << imuOutput_.qBW().z() << " " 
 << -imuOutput_.qBW().w() << std::endl;
 vio_result_file.close();

---

VI-ORB_SLAM

LearnVIORB 的代码地址
代码运行方法
轨迹生成
输出位置位于 System.cc

---

MSCKF

运行方法

roslaunch msckf_vio msckf_vio_euroc.launch 
rosrun rviz rviz -d ~/Project/msckf_vio_workspace/src/msckf_vio/rviz/rviz_euroc_config.rviz   //rviz显示模型
  rosbag play /home/wj/Downloads/dataset/EuRoC/ROS_bag/MH_05_difficult.bag

NOTE
The software does not run on EuRoC MH_01_easy.bag and MH_02_easy.bag. As explained in the README, the algorithm requires the sensor to start from staic in order to initialize the orientation and IMU bias. unfortunately, MH_01_easy.bag and MH_02_easy.bag do not have the initial static period.