boofcv.abst.geo.h

Class LeastSquaresHomography

java.lang.Object

boofcv.abst.geo.h.LeastSquaresHomography

All Implemented Interfaces:

RefineEpipolar, ModelFitter<org.ejml.data.DenseMatrix64F,AssociatedPair>

public class LeastSquaresHomography
extends java.lang.Object
implements RefineEpipolar

Improves upon the initial estimate of the Homography matrix by minimizing residuals.

Constructor Summary

Constructors

Constructor and Description
LeastSquaresHomography(double convergenceTol, int maxIterations, ModelObservationResidualN residuals)

Method Summary

Modifier and Type	Method and Description
boolean	fitModel(java.util.List<AssociatedPair> obs, org.ejml.data.DenseMatrix64F F, org.ejml.data.DenseMatrix64F refinedF) Fits a model to a set of points.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

LeastSquaresHomography

public LeastSquaresHomography(double convergenceTol,
                              int maxIterations,
                              ModelObservationResidualN residuals)

Method Detail

fitModel

public boolean fitModel(java.util.List<AssociatedPair> obs,
                        org.ejml.data.DenseMatrix64F F,
                        org.ejml.data.DenseMatrix64F refinedF)

Description copied from interface: ModelFitter

Fits a model to a set of points. Note that initial and found can be the same instance.

Specified by:

fitModel in interface ModelFitter<org.ejml.data.DenseMatrix64F,AssociatedPair>

Parameters:

obs - Set of points the model is being fit to.

F - Initial hypothesis

refinedF - The found model. Can be the same instance as initial.

Returns:

true if a model was found and false if one was not.