Image Component Library (ICL)
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines
Public Member Functions | Private Attributes
icl::markers::MultiCamFiducial Class Reference

Fiducial Type for the MuliCamFiducialDetector. More...

#include <MultiCamFiducial.h>

List of all members.

Public Member Functions

 MultiCamFiducial (MultiCamFiducialImpl *impl=0)
 Constructor.
bool isNull () const
 returns whether this is null-instance
int getCamsFound () const
 returns the number of cameras this fiducial was actually detected in
Fiducial operator[] (int idx)
 returns the underlying 2D fiducial detection result
const Fiducial operator[] (int idx) const
 returns the underlying 2D fiducial detection retult (const version)
geom::CameragetCamera (int idx)
 returns the idx-th associated camera
const geom::CameragetCamera (int idx) const
 returns the idx-th associated camera (const version)
int getID () const
 returns the associated marker ID
const math::FixedColVector
< float, 3 > & 
getCenter3D () const
 returns the 3D center of the marker
const math::FixedColVector
< float, 3 > & 
getOrientation3D () const
 returns the 3D orientation of the marker
const geom::MatgetPose3D () const
 returns the 6D-pose (in shape of a homogeneous transform) of the marker

Private Attributes

MultiCamFiducialImplimpl
 internal implementation

Detailed Description

Fiducial Type for the MuliCamFiducialDetector.

MultiCamFiducials provide much more robust 3D information if they a constructed from several camera views

The MultiCamFiducialDetector provides a generic interface to detect fiducials in a set of camera views

Memory management

The MultiCamFiducial is implemented as a wrapper around a single implementation pointer whose memory is managed by the parent MultiCamFiducialDetector instance. Therefore MultiCamFiducial instances can copied like objects

3D Pose estimation

For the pose estimation, two different algorithms are combined. In case of Fiducials, that were just detected in a single camera view, the CoplanarPointPoseEstimator is used. If more than one view is available, multicamera geometry is used to provide a much moch accurate estimation of the marker pose.


Constructor & Destructor Documentation

Constructor.


Member Function Documentation

returns the idx-th associated camera

Note, the camera can also be accessed from the idx-th Fiducial

returns the idx-th associated camera (const version)

returns the number of cameras this fiducial was actually detected in

The number of cameras can be a good indicator for the accuracy of the 3D position returned by getCenter3D(), getOrientation3D() and getPose3D()

returns the 3D center of the marker

See also:
3D Pose estimation

returns the associated marker ID

returns the 3D orientation of the marker

See also:
3D Pose estimation

returns the 6D-pose (in shape of a homogeneous transform) of the marker

See also:
3D Pose estimation
bool icl::markers::MultiCamFiducial::isNull ( ) const [inline]

returns whether this is null-instance

Fiducial icl::markers::MultiCamFiducial::operator[] ( int  idx)

returns the underlying 2D fiducial detection result

valid indices are [0,getCamsFound()-1]

const Fiducial icl::markers::MultiCamFiducial::operator[] ( int  idx) const

returns the underlying 2D fiducial detection retult (const version)


Member Data Documentation

internal implementation


The documentation for this class was generated from the following file:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines