Example #1
0
/*!
  Return the direct geometric model of the camera: fMc

  \param q : Articular position for pan and tilt axis.

  \return fMc, the homogeneous matrix corresponding to the direct geometric
  model of the camera. Describes the transformation between the robot reference
  frame (called fixed) and the camera frame.

*/
vpHomogeneousMatrix
vpPtu46::computeMGD (const vpColVector & q)
{
  vpHomogeneousMatrix fMc;

  computeMGD (q, fMc);

  return fMc;
}
Example #2
0
/*!
  Return the direct geometric model of the camera: fMc

  \warning Provided for compatibilty with previous versions. Use rather
  get_fMc(const vpColVector &).

  \param q : Articular position for pan and tilt axis.

  \return fMc, the homogeneous matrix corresponding to the direct geometric
  model of the camera. Describes the transformation between the robot reference
  frame (called fixed) and the camera frame.

  \sa get_fMc(const vpColVector &)
*/
vpHomogeneousMatrix
vpBiclops::computeMGD (const vpColVector & q) const
{
  vpHomogeneousMatrix fMc;

  computeMGD (q, fMc);

  return fMc;
}
Example #3
0
/*!
  Compute the direct geometric model of the camera in terms of pose vector.

  \param q : Articular position for pan and tilt axis.

  \param r : Pose vector corresponding to the transformation between the robot
  reference frame (called fixed) and the camera frame.

*/
void
vpPtu46::computeMGD (const vpColVector & q, vpPoseVector & r)
{
  vpHomogeneousMatrix fMc;

  computeMGD (q, fMc);
  r.buildFrom(fMc.inverse());

  return ;
}