请用c++写一段ros程序，gazebo仿真环境下有一个kinect深度相机，要求读取深度相机的深度图像和rgb图像，两者的话题名分别为/camera3/depth/image_raw和/camera3/rgb/image_raw，深度图片的编码格式为32FC1，读取图片并分别显示

好的，以下是示例代码，需要在ROS环境下使用：

#include <ros/ros.h>
#include <sensor_msgs/Image.h>
#include <sensor_msgs/image_encodings.h>
#include <cv_bridge/cv_bridge.h>
#include <opencv2/highgui/highgui.hpp>

void depthCallback(const sensor_msgs::ImageConstPtr& msg)
{
    if (msg->encoding != sensor_msgs::image_encodings::TYPE_32FC1) {
        ROS_ERROR("Depth image has unsupported encoding: %s", msg->encoding.c_str());
        return;
    }

    // Convert sensor_msgs/Image to OpenCV image
    cv_bridge::CvImagePtr cv_ptr;
    try {
        cv_ptr = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::TYPE_32FC1);
    } catch (cv_bridge::Exception& e) {
        ROS_ERROR("cv_bridge exception: %s", e.what());
        return;
    }

    // Display depth image
    cv::imshow("Depth Image", cv_ptr->image);
    cv::waitKey(1);
}

void rgbCallback(const sensor_msgs::ImageConstPtr& msg)
{
    // Convert sensor_msgs/Image to OpenCV image
    cv_bridge::CvImagePtr cv_ptr;
    try {
        cv_ptr = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::RGB8);
    } catch (cv_bridge::Exception& e) {
        ROS_ERROR("cv_bridge exception: %s", e.what());
        return;
    }

    // Display RGB image
    cv::imshow("RGB Image", cv_ptr->image);
    cv::waitKey(1);
}

int main(int argc, char** argv)
{
    ros::init(argc, argv, "kinect_image_viewer");
    ros::NodeHandle nh;

    // Subscribe to depth image topic
    ros::Subscriber depth_sub = nh.subscribe("/camera3/depth/image_raw", 1, depthCallback);

    // Subscribe to RGB image topic
    ros::Subscriber rgb_sub = nh.subscribe("/camera3/rgb/image_raw", 1, rgbCallback);

    // OpenCV window for showing images
    cv::namedWindow("Depth Image");
    cv::namedWindow("RGB Image");

    ros::spin();

    return 0;
}

这段代码会订阅深度图像和RGB图像的话题，并将其转换为OpenCV格式，然后使用OpenCV显示深度图像和RGB图像。注意，这里假设深度图像的编码格式为32FC1，RGB图像的编码格式为RGB8。如果实际情况不同，需要对代码进行相应修改。