link between detection algorithm and the knowledge base in Knowrob

edit

Hi all , I am working on an Roboearth application like Amigo who use roboearth to download object models to do a task

I am working now on detection package that detect the requested Object .

I tried re_object_detector_gui package that use the re_kinect_object_detector algorithm

BUT ,, this gui is manually by user choose or download the object model

1- How can i use the detect algorithm re_kinect_object_detector without this gui

by automatic detect the correspond object when i ask it for a cup for example this query search on knowrob data set and the result model detected by the detection algorithm ?

2- i wanna a sequence of how this process working like : first ,query from the user by .. Second : search in the knowledge base using ... third : if he found the model that match my request detect it by ... etec ...

thanks :)

Edited :

I made a simple talker node that publish an object model path in the re_kinect/model_path (std_msgs/String)

see the code :

#include "ros/ros.h"
#include "std_msgs/String.h"
#include <sstream>

int main(int argc, char **argv)
{

  ros::init(argc, argv, "talker");

  ros::NodeHandle n;

  ros::Publisher chatter_pub = n.advertise<std_msgs::String>("re_kinect/model_path", 1000);

  ros::Rate loop_rate(10);

  int count = 0;
  while (ros::ok())
  {

    std_msgs::String msg;

    std::stringstream ss;
    ss << "/home/turtlebot/Desktop/drink.b5ero_kinect" << count;
    msg.data = ss.str();

    ROS_INFO("%s", msg.data.c_str());
}

}

and rosmaked it with no error then :

1- run the openni launch

2- run the talker node , it returns the string of the model path

3- run the re_kinect node detector to detect

but it returns :

turtlebot@turtlebot-HP-Compaq-tc4400-EY609EA-ABB:~$ rosrun re_kinect_object_detector re_kinect
[ WARN] [1361999828.637136682]: no model files loaded!

how can it load the fixed model path i have entered all the time i run the detector ? where is the wrong ?

and that is the rxgraph for it :

thanks :)