mpiku / aria_client

Aria client for Pioneer robot.

Aria client for Pioneer robot

This piece of software along with aria_server is intended to help in integrating Pioneer robot and its camera with computer vision software which algorithms might be computationally expensive thus not usable on Pioneer's onboard computer.

Core of the software is robotManager class which is a container for four nested classes requestsHandler, keyHandlerMaster, cameraManager, steeringManager each being another smaller manager. These nested classes are responsible respectively for:

• requesting and receiving general information about the robot
• handling key strokes on client side
• receiving camera stream from robot and steering its camera
• steering the robot

Beginning

To reach properties and methods of nested classes one can use robotManager properties which are in fact instances of mentioned nested classes.

robotManager rManager( &argc, argv );

rManager.requests->methodOfClass_requestsHandler();
rManager.steering->methodOfClass_steeringManager();
rManager.camera->methodOfClass_cameraManager();
rManager.keyHandler->methodOfClass_keyHandlerMaster();

Verbose mode

To enable verbose mode in particular manager use enableVerboseMode() method.

rManager.camera->enableVerboseMode();

Steering Pioneer and its camera with client's keyboard

In order to activate robot's steering or camera's steering with keyboard use following methods.

• ← → ↑ ↓ - steering robot
• w s a d - camera pan / tilt
• r f - camer zoom in / out

rManager.keyHandler->startKeyMaster(); // Activate key strokes handling
rManager.steering->activateKeySteering(); // Activate robot's steering
rManager.camera->activateCameraSteering(); // Activate camera steering

Camera stream to OpenCV

To get live stream from Pioneer's camera to OpenCV object you can use following code.

cv::namedWindow( "Stream", CV_WINDOW_AUTOSIZE );
std::pair<unsigned char*, int> imageData;
while( true )
{
    imageData = rManager.camera->getSendVideoFrame();
    std::vector<unsigned char> buffer( imageData.first, imageData.first + imageData.second );
    cv::Mat image = imdecode(buffer, cv::IMREAD_ANYCOLOR);
    if(image.empty())
        return 0;
    cv::imshow("Stream", image);
    cv::waitKey( myStream->getSynchroTime_ums() / 1000 );
}