ROS2下编写orbbec相机C++ package并Rviz显示

视频讲解：

https://www.bilibili.com/video/BV1HUwSe3E6o/?vd_source=5ba34935b7845cd15c65ef62c64ba82f

开发环境：Ubuntu22.04 WSL
ROS2版本：humble

创建工作空间

mkdir ws_ros2
cd ws_ros2 
mkdir src
cd src

创建ros2 package模板，依赖cv_bridge，OpenCV，sensor_msgs

ros2 pkg create --build-type ament_cmake orbbec_cam_pkg --dependencies rclcpp sensor_msgs cv_bridge OpenCV

创建include和lib，include和lib的文件为orrbec sdk提取

cd orrbec_cam_pkg 
mkdir include lib

修改CMakeLists.txt，增加如下内容
file依赖的动态库
target链接
将动态库安装到工作空间的install的lib下

file(GLOB ORBBEC_LIBS "lib/x86_64/*")

target_link_libraries(image_publisher
    ${ORBBEC_LIBS}
)

install(DIRECTORY 
  lib/x86_64/
  DESTINATION lib/)

src下的image_publisher.cpp
发布消息类型为Image，设置发布定时器30ms，使用封装好的orbbec相机类获取cv图像，通过cv_bridge转换成msg

#include <rclcpp/rclcpp.hpp>
#include <sensor_msgs/msg/image.hpp>
#include <cv_bridge/cv_bridge.h>
#include <opencv2/opencv.hpp>
#include <thread>

#include "orbbec.h"

class ImagePublisher : public rclcpp::Node
{
public:
    ImagePublisher() : Node("image_publisher")
    {
        publisher_ = this->create_publisher<sensor_msgs::msg::Image>("rgb_image", 10);
        timer_ = this->create_wall_timer(std::chrono::milliseconds(30), std::bind(&ImagePublisher::timer_callback, this));
    }
    void camInit()
    {
        cam_.wait4Device();
        cam_.init();
        std::thread t = std::thread([&]() {
            cam_.run();
        });
        t.detach();
    }
private:
    void timer_callback()
    {
        // 创建一个简单的RGB图像（这里创建一个红色方块图像）
        // cv::Mat image(480, 640, CV_8UC3, cv::Scalar(0, 0, 255));
        cv::Mat image = cam_.getImg();

        // 将OpenCV图像转换为ROS 2图像消息
        sensor_msgs::msg::Image::SharedPtr msg = cv_bridge::CvImage(std_msgs::msg::Header(), "bgr8", image).toImageMsg();

        // 发布图像消息
        publisher_->publish(*msg);
        RCLCPP_INFO(this->get_logger(), "Publishing an RGB image");
    }

    rclcpp::Publisher<sensor_msgs::msg::Image>::SharedPtr publisher_;
    rclcpp::TimerBase::SharedPtr timer_;
    OrbbecCam cam_;
};

int main(int argc, char * argv[])
{
    rclcpp::init(argc, argv);
    auto node = std::make_shared<ImagePublisher>();
    node->camInit();
    rclcpp::spin(node);
    rclcpp::shutdown();
    return 0;
}

运行Rvi2，并添加Image显示

运行orbbec_cam_pkg程序，这里需要注意，需要sudo权限获取设备，所以先切到root，再source当前的ros2工作空间，不然会直接报错退出

sudo su
source install/setup.bash
ros2 run orbbec_cam_pkg image_publisher

可以看到已经获取到相机信息和发布出了数据，图像大小为480*640

切回到Rviz，将Image中的Topic选择为rgb_image，将History Policy选为System Default既可显示