由于slam_gmapping对里程计的严重依赖建图效果并不理想,尝试谷歌的Cartographer。再不使用OpenCV,PCL,g2o这些流量杀手的情况下能建出如此效果的地图还是让人惊叹的。简单说下原理:就是用Grid(2D/3D)的形式建地图;局部匹配直接建模成一个非线性优化问题,利用IMU提供一个比较靠谱的初值;后端用Graph来优化,用分支定界算法来加速;2D和3D的问题统一在一个框架下解决。(引用知乎--邵天兰)
0.安装所有依赖项
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-indigo-tf2-eigen libatlas-base-dev libsuitesparse-dev liblapack-dev1.首先安装ceres solver,选择的版本是1.11,路径随意
1. git clone https://github.com/hitcm/ceres-solver-1.11.0.git 2. cd ceres-solver-1.11.0/build 3. cmake .. 4. make –j 5. sudo make install
2.然后安装 cartographer,路径随意
1. git clone https://github.com/hitcm/cartographer.git 2 . cd cartographer/build 3. cmake .. -G Ninja 4. ninja 5. ninja test 6. sudo ninja install3.安装cartographer_ros。
谷歌官方提供的安装方法比较繁琐,我对原来的文件进行了少许的修改,核心代码不变,只是修改了编译文件 下载到catkin_ws下面的src文件夹下面 git clone https://github.com/hitcm/cartographer_ros.git 然后到catkin_ws下面运行catkin_make即可。
4.数据下载测试
2d数据,大概470M,用迅雷下载
https://storage.googleapis.com/cartographer-public-data/bags/backpack_2d/cartographer_paper_deutsches_museum.bag
然后运行launch文件即可。 roslaunch cartographer_ros demo_backpack_2d.launch bag_filename:=${HOME}/Downloads/cartographer_paper_deutsches_museum.bag
参考:https://www.cnblogs.com/hitcm/p/5939507.html