如何把cv::mat 转为compressedImage发送

您可以使用ROS的图像压缩和解压缩工具来将cv::Mat转换为CompressedImage并发送。具体过程如下：

1. 安装ROS 图像压缩和解压缩工具：`sudo apt-get install ros-<distro>-image-transport-plugins`

2. 导入必要的ROS头文件： `#include <ros/ros.h>` 和 `#include <sensor_msgs/CompressedImage.h>`

3. 声明ROS节点和发布器(注意指定话题名称和队列大小): `ros::NodeHandle nh; ros::Publisher pub = nh.advertise<sensor_msgs::CompressedImage>("compressed_image_topic", 1);`

4. 从cv::Mat创建一个ROS消息：`sensor_msgs::CompressedImage msg; msg.header.stamp = ros::Time::now(); msg.format = "jpeg"; cv::imencode(".jpg", input_image, buffer); msg.data.resize(buffer.size()); memcpy(&msg.data[0], &buffer[0], buffer.size());`

5. 将消息发布到话题： `pub.publish(msg);`

完整代码示例：

#include <ros/ros.h>
#include <sensor_msgs/CompressedImage.h>
#include <opencv2/opencv.hpp>

int main(int argc, char** argv)
{
    ros::init(argc, argv, "image_publisher");
    ros::NodeHandle nh;
    ros::Publisher pub = nh.advertise<sensor_msgs::CompressedImage>("compressed_image_topic", 1);

    cv::Mat input_image = cv::imread("input_image.jpg");

    sensor_msgs::CompressedImage msg;
    msg.header.stamp = ros::Time::now();
    msg.format = "jpeg";
    std::vector<uchar> buffer;
    cv::imencode(".jpg", input_image, buffer);
    msg.data.resize(buffer.size());
    memcpy(&msg.data[0], &buffer[0], buffer.size());

    pub.publish(msg);

    ros::spin();
}