I am trying to convert a cv::Mat into a sensor_msgs so that I can publish this in ROS.
my code is just like this :
while(ros::ok())
{
capture >> frame;
cv::imshow("Preview" , frame);
cv::waitKey(1);
//sensor_msgs::Image img_;
//fillImage(img_ , "rgb8" , frame.rows , frame.cols , 3 * frame.cols , frame);
//img_header.stamp = ros::Time::now();
//cv_bridge::CvImagePtr cv_ptr;
//cv_ptr->image = frame;
//image_pub_.publish(img_);
ros::spinOnce();
}
I have tried two potential solutions :
[1] using cv_bridge, CvImagePtr and toImageMsg(), but the CvImagePtr report
assert(px!0) error, which I guess means that I have to initialize CvImagePtr.
But I don't know how to initialize it;
[2] using fillImage and sensor_msgs::Image,
but the sixth parameter of fillImage has to be a void* instead of a Mat*
Hope anyone could help me !
Is there an efficient way to convert cv::Mat(or IplImage) to sensor_msgs ?
THX in advance !
Using image_transport: http://wiki.ros.org/image_transport/Tutorials/PublishingImages
Using cv_bridge interface and setting the header by hand
Use the following code
#include <sensor_msgs/Image.h>
#include <sensor_msgs/image_encodings.h>
#include <cv_bridge/cv_bridge.h>
cv::Mat img; // << image MUST be contained here
cv_bridge::CvImage img_bridge;
sensor_msgs::Image img_msg; // >> message to be sent
std_msgs::Header header; // empty header
header.seq = counter; // user defined counter
header.stamp = ros::Time::now(); // time
img_bridge = cv_bridge::CvImage(header, sensor_msgs::image_encodings::RGB8, img);
img_bridge.toImageMsg(img_msg); // from cv_bridge to sensor_msgs::Image
pub_img.publish(img_msg); // ros::Publisher pub_img = node.advertise<sensor_msgs::Image>("topic", queuesize);