基于Ubuntu14.04+ROS indigo环境LSD-SLAM的数据集测试成功

基于Ubuntu14.04+ROS indigo环境LSD-SLAM的数据集测试成功
作者：雨水/家辉  日期：2016-05-30  CSDN博客：http://blog.csdn.net/gobitan
摘要：LSD-SLAM是一个实时的单目SLAM. 它能实时地在电脑上直接(不使用关键点和特征)创建大规模的半稠密地图，甚至可以运行在智能手机上，如Android/iOS.

第一步：搭建Ubuntu14.04.4+ROS indigo基础环境
[1] 在http://releases.ubuntu.com/14.04/官网下载ubuntu-14.04.4-desktop-i386.iso安装包，大约1G。然后在VMware Player里安装好，可以采用VMPlayer的Easy Install自动化安装整个过程不需要干预。
备注：因为ROS包含rviz可视化环境，因此需要选择桌面版。

[2] 将安装源换成阿里云的，如何换请参考http://mirrors.aliyun.com/help/ubuntu;
备注：强烈建议将安装源换成国内的，否则安装的时候要么速度慢，要么可能连不上。

[3] 在Ubuntu14.04.4中安装ROS indigo，参考：http://wiki.ros.org/indigo/Installation/Ubuntu
dennis@ubuntu:~$ sudo apt-get update
dennis@ubuntu:~$ sudo sh -c 'echo "deb http://packages.ros.org/ros/ubuntu $(lsb_release -sc) main" > /etc/apt/sources.list.d/ros-latest.list'
dennis@ubuntu:~$ sudo apt-key adv --keyserver hkp://ha.pool.sks-keyservers.net --recv-key 0xB01FA116
dennis@ubuntu:~$ sudo apt-get update
dennis@ubuntu:~$ sudo apt-get install ros-indigo-desktop-full
备注：这里采用的完整安装模式.

dennis@ubuntu:~$ sudo rosdep init
dennis@ubuntu:~$ rosdep update

dennis@ubuntu:~$ echo "source /opt/ros/indigo/setup.bash" >> ~/.bashrc
dennis@ubuntu:~$ source ~/.bashrc

[4] 校验安装是否成功，输入roscore命令，如下：
dennis@ubuntu:~$ roscore
... logging to /home/dennis/.ros/log/9f2ace00-16b9-11e6-a10c-000c29b0df82/roslaunch-ubuntu-2078.log
Checking log directory for disk usage. This may take awhile.
Press Ctrl-C to interrupt
Done checking log file disk usage. Usage is <1GB.

started roslaunch server http://ubuntu:44364/ ros_comm version 1.11.19

SUMMARY
========

PARAMETERS
 * /rosdistro: indigo
 * /rosversion: 1.11.19

NODES

auto-starting new master
process[master]: started with pid [2090]
ROS_MASTER_URI=http://ubuntu:11311/

setting /run_id to 9f2ace00-16b9-11e6-a10c-000c29b0df82
process[rosout-1]: started with pid [2103]
started core service [/rosout]
[5] 安装ROS工具
dennis@ubuntu:~$ sudo apt-get install python-rosinstall

第二步：初始化ros编译环境及编译lsd_slam
dennis@ubuntu:~$ mkdir ~/rosbuild_ws
dennis@ubuntu:~$ cd ~/rosbuild_ws/
dennis@ubuntu:~/rosbuild_ws$ rosws init . /opt/ros/indigo
dennis@ubuntu:~/rosbuild_ws$ mkdir package_dir
dennis@ubuntu:~/rosbuild_ws$ rosws set ~/rosbuild_ws/package_dir -t .
dennis@ubuntu:~/rosbuild_ws$ echo "source ~/rosbuild_ws/setup.bash" >> ~/.bashrc
dennis@ubuntu:~/rosbuild_ws$ bash
dennis@ubuntu:~/rosbuild_ws$ cd package_dir/
dennis@ubuntu:~/rosbuild_ws/package_dir$
dennis@ubuntu:~/rosbuild_ws/package_dir$ sudo apt-get install ros-indigo-libg2o ros-indigo-cv-bridge liblapack-dev libblas-dev freeglut3-dev libqglviewer-dev libsuitesparse-dev libx11-dev
dennis@ubuntu:~/rosbuild_ws/package_dir$ git clone https://github.com/tum-vision/lsd_slam.git lsd_slam
dennis@ubuntu:~/rosbuild_ws/package_dir$ ls -lrt
total 4
drwxrwxr-x 5 dennis dennis 4096 May  8 03:30 lsd_slam
dennis@ubuntu:~/rosbuild_ws/package_dir$ rosmake lsd_slam

第三步：采用官方提供的数据集进行测试
备注：数据集下载请参考本文的参考资料[2].
测试方法一(采用image)：
dennis@ubuntu:~$ roscore
dennis@ubuntu:~$ rosrun lsd_slam_viewer viewer
dennis@ubuntu:~$ rosrun lsd_slam_core dataset_slam _files:=/home/dennis/LSD_room/images _hz:=0 _calib:=/home/dennis/LSD_room/cameraCalibration.cfg



测试方法二(采用bag文件)：
dennis@ubuntu:~$ roscore
dennis@ubuntu:~$ rosrun lsd_slam_viewer viewer
dennis@ubuntu:~$ rosrun lsd_slam_core live_slam image:=/image_raw camera_info:=/camera_info
dennis@ubuntu:~$ rosbag play LSD_room.bag






后记
这里只是做了简单的测试，获取直接的感官认识，后续将仔细阅读论文，阅读源代码，深入理解其实现原理及过程.

参考资料：
[1] https://github.com/tum-vision/lsd_slam   LSD_SLAM官网 
[2] http://vision.in.tum.de/research/vslam/lsdslam 德国慕尼黑工业大学(TUM)计算机视觉组 [3] http://wiki.ros.org/indigo/Installation/Ubuntu  在Ubuntu14.04.4上安装ROS Indigo
[4] http://releases.ubuntu.com/14.04/   官网下载ubuntu-14.04.4-desktop-i386.iso 
[5] http://wiki.ros.org/ROS/Tutorials 用于ROS安装成功之后的测试
[6] http://www.dev47apps.com/droidcam/linuxx/  将手机摄像头的视频传输到电脑上的工具