robwa's profile - activity

Hi, to open the kinect camera with OpenCV you need to use this:

VideoCapture capture( CV_CAP_OPENNI );

if you use the openni_kinect node to publish the data from the kinect, you could subscribe to one of the camera topics(/camera/rgb/image_color or /camera/rgb/image_mono). this would give you and image message which can be converted to an opencv iplimage using the cv_bridge package (http://www.ros.org/wiki/cv_bridge)

Duplicate http://answers.ros.org/question/1697/capture-for-kinect-camera - tfoote (Jul 26)

Thanks, that I have done, now I want to save the images I have listened to from to video using the video writer class, VideoCapture cap(0); // open the default camera However, I don't want to use the default camera (webcam), instead the kinect camera. I have tried -1, but still won't work. So, If someone has opened a capture using KINECT camera, that would be great if they shared with me how they did it. many thanks

Thanks, that I have done, now I want to save the images I have listened to from to video using the video writer class,

VideoCapture cap(0); // open the default camera

However, I don't want to use the default camera (webcam), instead the kinect camera. I have tried -1, but still won't work. So, If someone has opened a capture using KINECT camera, that would be great if they shared with me how they did it. many thanks

Hello There, I am trying to save a stream of images from KInect camera. MY first line of code is: CvCapture* capture = cvCaptureFromCAM(-1)

I passed -1 so that It can search for the camera because I don't know the index for KInect Camera. Can someone help me find out what it is or another method used in the past. Many thanks

I am trying to post a point in the world obtained using a kinect camera, and project it onto the image. However it gives me back point out of the normal range of a CV image, 640X480. I am using the function

  tf::Point pt = transform.getOrigin();
  cv::Point3d pt_cv(pt.x(), pt.y(), pt.z());
  cv::Point2d uv;
  cam_model_.project3dToPixel(pt_cv, uv);

Hi There, I am new at using ROS and Kinet. Right now I can do rostopic echo /tf and notice that there are frames being published. I however have another rospackage in which I have written a simple listener to listener to the frames published. I would like to know how to link openni_camera and my personal package, because right now it does not compile.

Thanks