RescueRobot-Movement_Detection

This Programm uses a rostopic, to track movement of a webcam. Around the detected motion there will be placed a colored bouning box. You can switch between the normal camera and the movement detection by pressing the button 'c' one time. If you repeat that, the image just switch back.

Tutorial

To set up this Package for your own Robot, there is one possibility.

1. Direct to your catkin-workspace.

cd catkin_ws/
cd src/

2. Clone this repository.

git clone https://github.com/CJT-Robotics/RescueRobot-Movement_Detection.git

(3.1) If you don't use our joy_relay [1], you have to manipulate our code. Remove this method.

void buttonCallback(const joy_relay::CJT& msg) {
  std::string action = msg.action;
  std::string device = msg.device;
  if(device.compare("camera") == 0)  {
    if(action.compare("movement_detection") == 0) {
      test = !test;
    }
  }
}

(3.2) Remove the btnSubscriber.

ros::Subscriber btnsub = nh.subscribe("/cjt/input", 1, buttonCallback);

4. Build your catkin_workspace.

cd ..
catkin_make

Dokumentation

The main method:

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

This initialize the ROS-Package. Here for the name you can choose your own name, but pay attention that this name could only used one time. If there is an other Package running with the same node-name, both will be crash.

image_transport::ImageTransport it(nh);

This will call the ImageTransport constructor.

image_transport::Subscriber imsub = it.subscribe("/cjt/image_raw", 1, imageCallback, image_transport::TransportHints("compressed"));

This subscribe the published image (Your Webcam or something else). You have to set your topic name to the one, you want to subscribe. Tipp to get the topic name: rostopic list imageCallback is the method to get the image and deal with it. If your topic supports compressed images, you need "image_transport::TransportHints("compressed")" because you can't subscribe the compressed topic directly.

cv::namedWindow("RoboCupGermany - CJT-Gymnasium - Movement Detection");

This is used for creating a window, later you put in your image.

ros::spin();

This is just for the subscriber and the threads.

The imageCallback method:

char c = cv::waitKey(30);

This function saves the user input (keypress). The time is formatted by milliseconds.

if(c == 'd') {
  test = !test;
}

Here it will be checked if the user input is the key 'd'. If its true the variable 'test' will be changed to the opposit position (true <==> false). So, this is the way how to change between seeing the movement detection or not. Also, you can switch the key there.

try {
//Space for code
} catch {
//Error message
}

In this try-catch-statement there are several things to be done.

frame = cv_bridge::toCvShare(msg, "bgr8")->image;

There the incoming picture/video will be converted to cv::Mat. This is for, to manipulate the image and create the movement detection system.

if (test == true) {
    detectMotion(frame);
}

Here, the check is running if the user selected to watch the movement detection or the plain image from the webcam.

cv::imshow("RoboCupGermany - CJT-Gymnasium - Movement Detection", frame);

This funtion is to paste the image into the created window.