本文共 6121 字，大约阅读时间需要 20 分钟。
Cartographer是google推出的基于激光雷达的SLAM算法。
Cartographer官方的编译安装步骤可以在上查看。但是在国内由于网络的原因，这个步骤并不适用。这里介绍它在国内网络环境下的安装和配置。
首先需要安装ROS，这里就不涉及，可以参考ROS的文档进行。ROS具体的版本没有影响，可以是Kinetic或者Melodic。
1 编译安装
Step 1 安装依赖
1) sudo apt-get update
2) sudo apt-get install -y python-wstool python-rosdep ninja-build
Step 2 创建目录carto_ws, 并初始化
$ mkdir carto_ws
$ cd carto_ws
$ wstool init src
输出如下：
robin@ubuntu:/home/bearpaw/carto_ws$ wstool init src
Writing /home/bearpaw/carto_ws/src/.rosinstall
update complete.
Step 3 新建一个文件/home/bearpaw/carto_ws/source.file
将 文件中的内容复制到此文件
Step 4 执行 wstool merge -t src /home/bearpaw/carto_ws/source.file
输出如下：
robin@ubuntu:/home/bearpaw/carto_ws$ wstool merge -t src /home/bearpaw/carto_ws/source.file
Performing actions:
Add new elements:
cartographer, cartographer_ros, ceres-solver
Config changed, maybe you need run wstool update to update SCM entries.
Overwriting /home/bearpaw/carto_ws/src/.rosinstall
update complete.
Step 5 执行完以上命令后，/home/bearpaw/carto_ws/src/.rosinstall中的内容为：
# THIS IS AN AUTOGENERATED FILE, LAST GENERATED USING wstool ON 2020-07-01
- git:
local-name: cartographer
uri:
version: 1.0.0
- git:
local-name: cartographer_ros
uri:
version: 1.0.0
- git:
local-name: ceres-solver
uri: git
version: 1.13.0
将最后一个git 替换为
Step 6 执行wstool update -t src
输出如下：
robin@ubuntu:/home/bearpaw/carto_ws$ wstool update -t src
[cartographer] Fetching https://github.com/cartographer-project/cartographer.git (version 1.0.0) to /home/bearpaw/carto_ws/src/cartographer
Cloning into '/home/bearpaw/carto_ws/src/cartographer'...
remote: Enumerating objects: 21, done.
remote: Counting objects: 100% (21/21), done.
remote: Compressing objects: 100% (19/19), done.
remote: Total 14043 (delta 5), reused 8 (delta 2), pack-reused 14022 Receiving objects: 100% (14043/14043), 5.44 MiB | 10.00 KiB/s, done.
Resolving deltas: 100% (11187/11187), done.
[cartographer] Done.
[cartographer_ros] Fetching https://github.com/cartographer-project/cartographer_ros.git (version 1.0.0) to /home/bearpaw/carto_ws/src/cartographer_ros
Cloning into '/home/bearpaw/carto_ws/src/cartographer_ros'...
remote: Enumerating objects: 13, done. remote: Counting objects: 100% (13/13), done.
remote: Compressing objects: 100% (12/12), done.
remote: Total 4302 (delta 2), reused 1 (delta 1), pack-reused 4289
Receiving objects: 100% (4302/4302), 4.09 MiB | 28.00 KiB/s, done.
Resolving deltas: 100% (3140/3140), done.
[cartographer_ros] Done.
[ceres-solver] Fetching https://github.com/ceres-solver/ceres-solver.git (version 1.13.0) to /home/bearpaw/carto_ws/src/ceres-solver
Cloning into '/home/bearpaw/carto_ws/src/ceres-solver'...
remote: Enumerating objects: 8, done.
remote: Counting objects: 100% (8/8), done.
remote: Compressing objects: 100% (8/8), done.
remote: Total 14355 (delta 0), reused 2 (delta 0), pack-reused 14347
Receiving objects: 100% (14355/14355), 13.25 MiB | 30.00 KiB/s, done.
Resolving deltas: 100% (9215/9215), done.
[ceres-solver] Done.
Step 7 执行 src/cartographer/scripts/install_proto3.sh
下载编译合适的proto3
输出如下：
robin@ubuntu:/home/bearpaw/carto_ws$ src/cartographer/scripts/install_proto3.sh
VERSION="v3.4.1"
# Build and install proto3.
git clone https://github.com/google/protobuf.git
Cloning into 'protobuf'...
remote: Enumerating objects: 12, done.
remote: Counting objects: 100% (12/12), done.
remote: Compressing objects: 100% (11/11), done.
注意：
1）第6和7步，实际上是下载几个github的包，如果网速慢，会经常下载失败，这时，可以直接从github上download zip包然后解压到相应的目录即可。
github地址为：
v1.0.0
v1.0.0
https://github.com/ceres-solver/ceres-solver v1.13.0
v3.4.1
目录结构如下布置：
robin@ubuntu:/home/bearpaw/carto_ws$ ls
protobuf src
robin@ubuntu:/home/bearpaw/carto_ws$ ls ./src/
cartographer cartographer_ros ceres-solver
2）protobuf下载完之后，将src/cartographer/scripts/install_proto3.sh的内容修改如下：
$cat install_proto3.sh
#!/bin/sh
set -o errexit
set -o verbose
VERSION="v3.4.1"
# Build and install proto3.
cd protobuf
mkdir build
cd build
cmake -G Ninja \
-DCMAKE_POSITION_INDEPENDENT_CODE=ON \
-DCMAKE_BUILD_TYPE=Release \
-Dprotobuf_BUILD_TESTS=OFF \
../cmake
ninja
sudo ninja install
Step 8 执行下面三个命令，安装依赖
sudo rosdep init （以前执行过此命令，会报错， 忽略即可）
rosdep update
rosdep install --from-paths src --ignore-src --rosdistro=${ROS_DISTRO} -y
Step 9 编译安装
> catkin_make_isolated --install --use-ninja
2配置
2.1 配置文件
安装之后，配置文件位于/home/bearpaw/carto_ws/install_isolated/share/cartographer_ros/configuration_files
在此目录创建自己的配置文件， 我的名为bearpaw.lua,内容如下：
$ cat bearpaw.lua
-- Copyright 2016 The Cartographer Authors
include "map_builder.lua"
include "trajectory_builder.lua"
options = {
map_builder = MAP_BUILDER,
trajectory_builder = TRAJECTORY_BUILDER,
map_frame = "map",
tracking_frame = "base_footprint",
published_frame = "base_footprint",
odom_frame = "odom_combined",
provide_odom_frame = true,
publish_frame_projected_to_2d = false,
use_odometry = false,
use_nav_sat = false,
use_landmarks = false,
num_laser_scans = 1,
num_multi_echo_laser_scans = 0,
num_subdivisions_per_laser_scan = 1,
num_point_clouds = 0,
lookup_transform_timeout_sec = 0.2,
submap_publish_period_sec = 0.3,
pose_publish_period_sec = 5e-3,
trajectory_publish_period_sec = 30e-3,
rangefinder_sampling_ratio = 1.,
fixed_frame_pose_sampling_ratio = 1.,
odometry_sampling_ratio = 1.,
imu_sampling_ratio = 1.,
landmarks_sampling_ratio = 1.,
}
MAP_BUILDER.use_trajectory_builder_2d = true
TRAJECTORY_BUILDER_2D.use_online_correlative_scan_matching = true
TRAJECTORY_BUILDER_2D.use_imu_data = false
TRAJECTORY_BUILDER_2D.real_time_correlative_scan_matcher.linear_search_window = 0.15
TRAJECTORY_BUILDER_2D.real_time_correlative_scan_matcher.angular_search_window = math.rad(35.)
POSE_GRAPH.optimization_problem.huber_scale = 1e2
标红色的内容需要根据自己的实际情况配置。
2.2 Launch file
配置文件之后，编写launch file，内容如下：
$ cat demo_bearpaw.launch
<launch>
<node name="cartographer_node" pkg="cartographer_ros"
type="cartographer_node" args="
-configuration_directory
$(find cartographer_ros)/configuration_files
-configuration_basename bearpaw.lua"
output="screen">
</node>
<node name="cartographer_occupancy_grid_node" pkg="cartographer_ros"
type="cartographer_occupancy_grid_node" args="-resolution 0.05" />
</launch>
下面就可以用这个launch 文件启动cartographer了。
转载地址：https://blog.csdn.net/lclfans1983/article/details/107256133 如侵犯您的版权，请留言回复原文章的地址，我们会给您删除此文章，给您带来不便请您谅解！
发表评论
最新留言
关于作者
