ubuntu16.04 + ros-kinetic 配置cartographer

其实一直以来都感觉纯视觉SLAM很难落地产品，所以一直在找机会学习激光slam,之前也在深蓝学院上买了一个激光salm的课程，惭愧，至今也没开始学呢，年底之前，我想工作之余研究一下激光slam和ros，我感觉这两个东西对我来说很重要，所以这两天先抽时间把google的cartographer工程配置了一下，期间也遇到了一些问题，不过我逐一解决，最终配置成功。接下来就开始研究激光slam技术，以及cartographer的代码。希望能帮上和我一起想开始学习cartographer的小伙伴，如果有什么问题，可以随时想我提问，大家一起学习。

更新：注意：我之前安装了conda,conda里面的python版本是3.7的，我在安装cartographer_ros的时候没有关闭conda所以出现了错误，因为ros只能在python2.7版本下才能正常工作，所以我就conda deactivate base.这样就恢复到了原来的python2.7的版本。

参考文献如下

https://google-cartographer.readthedocs.io/en/latest/

https://blog.csdn.net/xmy306538517/article/details/81455625

https://www.cnblogs.com/hitcm/p/5939507.html

step1：安装一些依赖库

# Install the required libraries that are available as debs.
sudo apt-get update
sudo apt-get install -y \clang \cmake \g++ \git \google-mock \libboost-all-dev \libcairo2-dev \libcurl4-openssl-dev \libeigen3-dev \libgflags-dev \libgoogle-glog-dev \liblua5.2-dev \libsuitesparse-dev \lsb-release \ninja-build \stow# Install Ceres Solver and Protocol Buffers support if available.
# No need to build it ourselves.
if [[ "$(lsb_release -sc)" = "focal" || "$(lsb_release -sc)" = "buster" ]]
thensudo apt-get install -y python3-sphinx libgmock-dev libceres-dev protobuf-compiler
elsesudo apt-get install -y python-sphinxif [[ "$(lsb_release -sc)" = "bionic" ]]thensudo apt-get install -y libceres-devfi
fi

step2：安装abseil，注意，在最后一步执行sudo stow absl之前要先执行sudo apt-get install stow安装stow

git clone https://github.com/abseil/abseil-cpp.git
cd abseil-cpp
git checkout d902eb869bcfacc1bad14933ed9af4bed006d481
mkdir build
cd build
cmake -G Ninja \-DCMAKE_BUILD_TYPE=Release \-DCMAKE_POSITION_INDEPENDENT_CODE=ON \-DCMAKE_INSTALL_PREFIX=/usr/local/stow/absl \..
ninja
sudo ninja install
cd /usr/local/stow
sudo stow absl

step3: 这个下载链接是google的资源，如果你没有设置全局翻墙，在terminal不可翻墙，就不能下载成功，解决方法是在github里搜索ceres-solver,然后采用git clone的方式下载，如果在网页上直接Download下载下来的工程中灭有.git文件，无法完成git checkout tags/${VERSION}

VERSION="1.13.0"# Build and install Ceres.
git clone https://ceres-solver.googlesource.com/ceres-solver
cd ceres-solver
git checkout tags/${VERSION}
mkdir build
cd build
cmake .. -G Ninja -DCXX11=ON
ninja
CTEST_OUTPUT_ON_FAILURE=1 ninja test
sudo ninja install

step4: 安装protobuf.使用下面的下载链接下载时如果一直出现early EOF的超时错误时，就使用

git clone --depth=1 https://github.com/google/protobuf.git 进行下载，--depth=1是约束了只下载最新git log的版本。

VERSION="v3.4.1"# Build and install proto3.
git clone https://github.com/google/protobuf.git
cd protobuf
git checkout tags/${VERSION}
mkdir build
cd build
cmake -G Ninja \-DCMAKE_POSITION_INDEPENDENT_CODE=ON \-DCMAKE_BUILD_TYPE=Release \-Dprotobuf_BUILD_TESTS=OFF \../cmake
ninja
sudo ninja install

step5: 安装cartographer

# Build and install Cartographer.git clone git@github.com:cartographer-project/cartographer.git
cd cartographer
mkdir build
cd build
cmake .. -G Ninja
ninja
CTEST_OUTPUT_ON_FAILURE=1 ninja test
sudo ninja install

step6:安装cartographer_ros

注意：创建ros工程的workspace时，根据你自己的路径来就可以，还有在将工程的devel中的setup.bash的路径写入到.bashrc中时也要注意你的路径。这样做的好处你，你新打开一个窗口进行操作时，不需要重新执行source YOUR_PATH/devel/setup.bash了就。

sudo apt-get install -y python-wstool python-rosdep ninja-buildmkdir -p /code/ROS/catkin_ws_cartographer/src
cd /code/ROS/catkin_ws_cartographer/src
git clone --recursive https://github.com/googlecartographer/cartographer_ros.gitcd /code/ROS/catkin_ws_cartographer
catkin_make_isolated --install --use-ninja -Dcartographer_DIR="/usr/local/share/cartographer/"
source /code/ROS/catkin_ws_cartographer/devel_isolated/setup.bash
echo "source /code/ROS/catkin_ws_cartographer/devel_isolated/setup.bash" >> ~/.bashrc

step7:下载数据集

2d数据集：

https://storage.googleapis.com/cartographer-public-data/bags/backpack_2d/cartographer_paper_deutsches_museum.bag

3d数据集：

https://storage.googleapis.com/cartographer-public-data/bags/backpack_3d/b3-2016-04-05-14-14-00.bag

step8: 运行程序

注意：修改成你自己的存放数据集的路径。

roslaunch cartographer_ros demo_backpack_2d.launch bag_filename:=${HOME}/Downloads/cartographer_paper_deutsches_museum.bag
roslaunch cartographer_ros demo_backpack_3d.launch bag_filename:=${HOME}/Downloads/cartographer_3d_deutsches_museum.bag

step9: 测试结果展示