1.cartorgrapher介绍

无需过多介绍，大名鼎鼎google开源的slam算法

2.硬件需求

先贴段官方的

64-bit, modern CPU (e.g. 3rd generation i7)
16 GB RAM
Ubuntu 14.04 (Trusty) and 16.04 (Xenial)
gcc version 4.8.4 and 5.4.0

当然越高端越好无论CPU和内存，经测试树莓派是没法玩了，测试了2款RK3399(Ubuntu16)设备可以跑起来

3.环境准备与编译

3.1安装依赖包

sudo apt-get install -y google-mock libboost-all-dev  libeigen3-dev libgflags-dev libgoogle-glog-dev liblua5.2-dev libprotobuf-dev  libsuitesparse-dev libwebp-dev ninja-build protobuf-compiler python-sphinx  ros-kinetic-tf2-eigen libatlas-base-dev libsuitesparse-dev liblapack-dev ros-kinetic-eigen-conversions

3.2编译

• 编译ceres-solver

cd pibot_ros
tar jxvf cartographer.tar.bz2
cd ~pibot_ros/cartographer/ceres-solver/build
cmake ..
make –j4  # 这里如果出错（见下图），可以多次执行
sudo make install

make –j4报错

• 继续cartographer

cd ~pibot_ros/cartographer/cartographer/build
cmake .. -G Ninja
ninja
ninja test
sudo ninja install

• 编译cartorgrapher ros 包

cd ~pibot_ros/cartographer/cartographer_ros
catkin_make

4.测试cartographer

cartograper不依赖里程计，所以不需要发出odom tf,这里我们启动bringup_without_odom.launch

roslaunch pibot_bringup bringup_without_odom.launch

另外一个终端

source ~pibot_ros/cartographer/cartographer_ros/devel/setup.bash
roslaunch pibot_navigation cartographer.launch

另外一个终端，通过控制小车就可以cartographer建图了

roslaunch pibot keyboard.launch

我们先看下tf tree,虚拟机中rosrun rqt_tf_tree rqt_tf_tree可以看到一幅支持的tf tree了

5.查看建图

可是说好的建图的，图呢，没图说个jb.jpg,同样我们在rviz显示建图结果

roslaunch pibot_navigation view_cartographer.launch

6.疑问

对于4.测试cartographer中
有的同学要问为什么不把这个2个写到一个launch,这个问题问的好，博主试着写到一起，启动不正常 ，还不知道具体原因，有兴趣的可以一起探讨

附上之前的launch文件

<launch>
  <include file="$(find pibot_bringup)/launch/robot_without_odom.launch"/>
  <include file="$(find cartographer_ros)/launch/lidar_2d.launch"/>
</launch>