文章目录

1.安装准备

这里假设你已经装好了ROS2以及Moveit2（都用二进制安装就行，不用从源码安转），没有安装好的，可以按照鱼香ROS的教程安装，两三行命令就搞定了。
我的ROS2版本为humble，请根据你使用的实际版本替换。
安装pcl

sudo apt install libpcl-dev
sudo apt-get install ros-humble-pcl*

安装octomap相关库

sudo apt-get install ros-humble-octomap*

下面这个也安装一下

sudo apt-get install ros-humble-moveit-ros-perception

2.#include需要的头文件

由于涉及到的功能比较繁琐，因此需要包含进来的头文件也很多。之前忘记把头文件给贴出来了，导致可能很多同学拷贝了后面的代码进行编译时，各种报错。
现在把所可能用到的头文件都贴出来。
编译时的头文件路径（以你自己的为准）：

# 一系列头文件所在的文件夹
INCLUDEPATH += /opt/ros/humble/include\
/opt/ros/humble/include/rclcpp \
/opt/ros/humble/include/moveit \
/opt/ros/humble/include/rcl \
/opt/ros/humble/include/rcutils \
/opt/ros/humble/include/rmw \
/opt/ros/humble/include/rcl_yaml_param_parser \
/opt/ros/humble/include/rosidl_runtime_c \
/opt/ros/humble/include/rosidl_typesupport_interface \
/opt/ros/humble/include/rcpputils \
/opt/ros/humble/include/builtin_interfaces \
/opt/ros/humble/include/rosidl_runtime_cpp \
/opt/ros/humble/include/tracetools \
/opt/ros/humble/include/rcl_interfaces \
/opt/ros/humble/include/libstatistics_collector \
/opt/ros/humble/include/statistics_msgs \
/opt/ros/humble/include/moveit_msgs \
/opt/ros/humble/include/tf2_ros \
/opt/ros/humble/include/tf2 \
/opt/ros/humble/include/tf2_msgs \
/opt/ros/humble/include/geometry_msgs \
/opt/ros/humble/include/std_msgs \
/opt/ros/humble/include/sensor_msgs \
/opt/ros/humble/include/visualization_msgs \
/opt/ros/humble/include/shape_msgs \
/opt/ros/humble/include/object_recognition_msgs \
/opt/ros/humble/include/octomap_msgs \
/opt/ros/humble/include/unique_identifier_msgs \
/opt/ros/humble/include/action_msgs \
/opt/ros/humble/include/rclcpp_action \
/opt/ros/humble/include/rcl_action \
/opt/ros/humble/include/graph_msgs \
/opt/ros/humble/include/tf2_eigen \
/opt/ros/humble/include/tf2_geometry_msgs \
/opt/ros/humble/include/urdf \
/opt/ros/humble/include/trajectory_msgs \
/usr/include/eigen3

# 库文件
LIBS += -L/opt/ros/humble/lib
LIBS += /opt/ros/humble/lib/*.so
LIBS += -L/opt/ros/humble/lib/x86_64-linux-gnu \
-loctomap

INCLUDEPATH += /usr/include/pcl-1.12
LIBS += /usr/lib/x86_64-linux-gnu/libpcl_*

代码中需要的头文件：

#include <moveit/occupancy_map_monitor/occupancy_map_monitor.h>
#include <moveit/planning_scene_monitor/planning_scene_monitor.h>
#include <moveit/move_group_interface/move_group_interface.h>
#include <moveit_visual_tools/moveit_visual_tools.h>
#include <moveit/planning_scene_interface/planning_scene_interface.h>
#include <moveit/kinematic_constraints/utils.h>
#include <moveit_msgs/msg/constraints.hpp>

#include <tf2_eigen/tf2_eigen.hpp>
#include <tf2_geometry_msgs/tf2_geometry_msgs.hpp>
#include <tf2/LinearMath/Quaternion.h>
#include <tf2/LinearMath/Matrix3x3.h>

#include <octomap/OcTree.h>
#include <octomap/octomap.h>
#include <octomap_msgs/conversions.h>
#include <octomap_msgs/msg/octomap.hpp>
#include <octomap_msgs/conversions.h>

#include <pcl/pcl_base.h>
#include <pcl/PCLPointCloud2.h>
#include <pcl/point_cloud.h>
#include <pcl/point_types.h>
#include <pcl/point_types_conversion.h>
#include <pcl/conversions.h>
#include <pcl/common/transforms.h>
#include <pcl/io/pcd_io.h>

3.bt文件读取

octomap信息可能来自话题（由设备或者其他节点发布），或者从bt文件中读取。这里出于便于演示等原因，选择从文件中读取。文件读取时要注意使用的bt文件是txt还是bin的，要选择对应的读取方式，具体请查看代码。
这个octomap的功能还是用之前做的机械手环境来演示。
以下所使用到的【geb079.bt】文件来自于octomap的github。而【octomap.bt】文件是通过以下函数创建：

void createOctoMap()
{
    // 创建OcTree对象，并设置分辨率为0.05
    octomap::OcTree* tree = new octomap::OcTree(0.05);

    // 设置障碍物的大小和位置
    octomap::point3d origin(0, 0, 0);
    octomap::point3d obstacle(0, 0, 0); // 1 1 1
    double size = 0.2;

    // 在OcTree中添加一个立方体障碍物
    tree->updateNode(octomap::point3d(origin.x() + obstacle.x() - size, origin.y() + obstacle.y() - size, origin.z() + obstacle.z() - size), true);
    tree->updateNode(octomap::point3d(origin.x() + obstacle.x() + size, origin.y() + obstacle.y() - size, origin.z() + obstacle.z() - size), true);
    tree->updateNode(octomap::point3d(origin.x() + obstacle.x() + size, origin.y() + obstacle.y() + size, origin.z() + obstacle.z() - size), true);
    tree->updateNode(octomap::point3d(origin.x() + obstacle.x() - size, origin.y() + obstacle.y() + size, origin.z() + obstacle.z() - size), true);
    tree->updateNode(octomap::point3d(origin.x() + obstacle.x() - size, origin.y() + obstacle.y() - size, origin.z() + obstacle.z() + size), true);
    tree->updateNode(octomap::point3d(origin.x() + obstacle.x() + size, origin.y() + obstacle.y() - size, origin.z() + obstacle.z() + size), true);
    tree->updateNode(octomap::point3d(origin.x() + obstacle.x() + size, origin.y() + obstacle.y() + size, origin.z() + obstacle.z() + size), true);
    tree->updateNode(octomap::point3d(origin.x() + obstacle.x() - size, origin.y() + obstacle.y() + size, origin.z() + obstacle.z() + size), true);

    // 将OcTree保存为octomap.bt文件
    std::string filename = "octomap.bt";
    auto ret = tree->writeBinary(filename);
//    auto ret = tree->write(filename);

    qDebug() << "write octomap.bt:" << ret;
    // 删除OcTree对象
    delete tree;
}

假如要将读取、接收到的octomap通过代码的方式“输出”到moveit中，让其参与碰撞检测、路径规划，一般有两种方式：

3.1通过moveit_msgs::msg::CollisionObject加入场景中

这种方式，是通过 octomap–》pointCloud–》mesh 的方式进行的，比较麻烦。而且在pointCloud–》mesh这一步遇到了一些问题，转出来的是面片，而非实体mesh。后面再研究研究

// 这个函数还是有问题的，弄出来的是面片，而不是mesh，后续还需要改进
void pclPointCloudToShapeMsgsMesh(const pcl::PointCloud<pcl::PointXYZ> *cloud,
                                  shape_msgs::msg::Mesh &mesh)
{
    // 创建三角形索引数组
    pcl::PolygonMesh triangles;
    pcl::Vertices vertices;

    for (size_t i = 0; i < cloud->size() - 2; ++i)
    {
        vertices.vertices.clear();
        vertices.vertices.push_back(i);
        vertices.vertices.push_back(i + 1);
        vertices.vertices.push_back(i + 2);
        triangles.polygons.push_back(vertices);
    }

    // 拷贝点云数据到Mesh消息中
    mesh.vertices.resize(cloud->size());
    for (int i = 0; i < cloud->size(); ++i)
    {
        geometry_msgs::msg::Point p;
        p.x = (*cloud)[i].x;
        p.y = (*cloud)[i].y;
        p.z = (*cloud)[i].z;
        mesh.vertices[i] = p;
    }

    // 拷贝三角形索引数据到Mesh消息中
    mesh.triangles.resize(triangles.polygons.size());
    for (int i = 0; i < triangles.polygons.size(); ++i)
    {
        shape_msgs::msg::MeshTriangle t;
        t.vertex_indices[0] = triangles.polygons[i].vertices[0];
        t.vertex_indices[1] = triangles.polygons[i].vertices[1];
        t.vertex_indices[2] = triangles.polygons[i].vertices[2];
        mesh.triangles[i] = t;
    }
}

// 将octoMap作为碰撞物体导入到moveit中
void appendCollisionObject(QString octoFile)
{
    auto const node = std::make_shared<rclcpp::Node>(
                "hello_moveit",
                rclcpp::NodeOptions().automatically_declare_parameters_from_overrides(true));

    //创建 MoveIt MoveGroup Interface，用于连接、获取MoveGroup的信息
    using moveit::planning_interface::MoveGroupInterface;
    auto move_group_interface = MoveGroupInterface(node, "my_group"); // 估计要依附一个节点，才能进行interface的创建

    // 将碰撞物体加到scene中
    moveit::planning_interface::PlanningSceneInterface planning_scene_interface;

    auto frameId = move_group_interface.getPlanningFrame();
    qDebug() << "frame id:" << frameId.data();
    {
        std::shared_ptr<octomap::OcTree> octree = std::make_shared<octomap::OcTree>(octomap::OcTree(octoFile.toStdString()));
        if (octree == nullptr) {
          throw std::runtime_error("Could not read Octomap data from file " + octoFile.toStdString());
        }

        // octTree转点云
        pcl::PointCloud<pcl::PointXYZ> cloud;
        // Traverse all the leaf nodes and add them to the point cloud if they are occupied
        for (auto it = octree->begin_leafs(); it != octree->end_leafs(); ++it) {
            if (octree->isNodeOccupied(*it)) {
                pcl::PointXYZ pt;
                pt.x = it.getX();
                pt.y = it.getY();
                pt.z = it.getZ();
                cloud.points.push_back(pt);
            }
        }

        // 点云转mesh
        shape_msgs::msg::Mesh output_mesh;
        pclPointCloudToShapeMsgsMesh(&cloud, output_mesh);

        geometry_msgs::msg::Pose octomap_pose;
        octomap_pose.orientation.w = 1.0;
        octomap_pose.position.x = 0.2;
        octomap_pose.position.y = 0.2;
        octomap_pose.position.z = 0.2;

        // 直接往octomap_object里面塞mesh
        moveit_msgs::msg::CollisionObject octomap_object;
        octomap_object.id = "my_octomap_object";
        octomap_object.header.frame_id = frameId;
        octomap_object.meshes.push_back(output_mesh);
        octomap_object.mesh_poses.push_back(octomap_pose);
        octomap_object.operation = moveit_msgs::msg::CollisionObject::ADD;


        // 将碰撞对象添加到collision_objects中，然后传递给applyCollisionObjects函数
        std::vector<moveit_msgs::msg::CollisionObject> collision_objects;
        collision_objects.push_back(octomap_object);
        qDebug() << "applyCollisionObjects" <<  planning_scene_interface.applyCollisionObjects(collision_objects);
    }
}

3.2通过moveit_msgs::msg::PlanningScene加入（替换）场景

目前我使用的是这种，效果良好。不过可能会存在一些冲突，等后面研究得更加透彻之后再修改吧。

void appendScene(QString octoFile)
{
    auto const node = std::make_shared<rclcpp::Node>(
                "hello_moveit1",
                rclcpp::NodeOptions().automatically_declare_parameters_from_overrides(true));


    //创建 MoveIt MoveGroup Interface，用于连接、获取MoveGroup的信息
    using moveit::planning_interface::MoveGroupInterface;
    auto move_group_interface = MoveGroupInterface(node, "my_group"); // 估计要依附一个节点，才能进行interface的创建
    auto frameId = move_group_interface.getPlanningFrame();

    // Load an OctoMap from file
    octomap::OcTree* octree = new octomap::OcTree(octoFile.toStdString());

    // 将Octomap数据转换为ROS2消息
    octomap_msgs::msg::Octomap octomap_msg;
    octomap_msgs::binaryMapToMsg(*octree, octomap_msg);

    // 创建PlanningSceneMonitor实例, 用于获取 AllowedCollisionMatrix
    auto planning_scene_monitor = std::make_shared<planning_scene_monitor::PlanningSceneMonitor>(node,
        "robot_description");
    collision_detection::AllowedCollisionMatrix acm = planning_scene_monitor->getPlanningScene()->getAllowedCollisionMatrix();
//    // 假设我们需要允许"obstacle_1"和"obstacle_2"之间的碰撞
//    acm.setEntry("obstacle_1", "obstacle_2", true);

    // 将 collision_detection::AllowedCollisionMatrix 转成 moveit_msgs::msg::AllowedCollisionMatrix
    moveit_msgs::msg::AllowedCollisionMatrix acm_msg;
    acm.getMessage(acm_msg);

    moveit_msgs::msg::PlanningScene planning_scene_msg;
    planning_scene_msg.world.octomap.header.frame_id = frameId;  // Replace "map" with your frame ID
    planning_scene_msg.world.octomap.origin.orientation.w = 1.0;
    planning_scene_msg.world.octomap.octomap = octomap_msg;
    planning_scene_msg.allowed_collision_matrix = acm_msg;

    // 将PlanningScene 加（替换）到moveit中
    moveit::planning_interface::PlanningSceneInterface planning_scene_interface;
    planning_scene_interface.applyPlanningScene(planning_scene_msg);
}

假如想从点云转octomap，也非常简单。转成octomap之后，也就可以通过上面的方式导入moveIt中了。函数参考这里。

// pointCloud转成octomap
void pointCloudToOctoMap(const pcl::PointCloud<pcl::PointXYZ> *cloud,
                                  octomap_msgs::msg::Octomap  &octoMap)
{
    // create octomap
    octomap::OcTree tree(0.05); // set resolution of the octomap
    for (const auto& point : cloud->points)
    {
      tree.updateNode(octomap::point3d(point.x, point.y, point.z), true);
    }
    tree.updateInnerOccupancy(); // update inner node occupancy

   // convert octomap to octomap_msgs::msg::Octomap
    octomap_msgs::binaryMapToMsg(tree, octoMap);
}

4.效果演示

视频演示可以看这里：【MoveIt2中导入octomap】
在这里插入图片描述

参考
https://blog.csdn.net/qq_27865227/article/details/125002311

# 计算机视觉 # 机器人 # opencv
Logo
开放原子开发者工作坊

开放原子开发者工作坊旨在鼓励更多人参与开源活动，与志同道合的开发者们相互交流开发经验、分享开发心得、获取前沿技术趋势。工作坊有多种形式的开发者活动，如meetup、训练营等，主打技术交流，干货满满，真诚地邀请各位开发者共同参与！

更多推荐

第二届开放原子大赛首批创新成果集结武汉，诚邀广大开发者共鉴开源技术盛宴

第二届开放原子大赛首批创新成果集结武汉，诚邀广大开发者共鉴开源技术盛宴

avatar 开放原子开发者工作坊

诚邀报名 | 开源基础设施能力建设分论坛：打造开源生态的“心脏”

诚邀报名 | 开源基础设施能力建设分论坛：打造开源生态的“心脏”

avatar 开放原子开发者工作坊

诚邀报名 | 编程语言分论坛：AI时代的技术革新与开源实践

诚邀报名 | 编程语言分论坛：AI时代的技术革新与开源实践

avatar 开放原子开发者工作坊