org.openimaj.math.model.fit

Class RANSAC.ProbabilisticMinInliersStoppingCondition

java.lang.Object

org.openimaj.math.model.fit.RANSAC.ProbabilisticMinInliersStoppingCondition

All Implemented Interfaces:

RANSAC.StoppingCondition

Enclosing class:

RANSAC<I,D,M extends EstimatableModel<I,D>>

public static class RANSAC.ProbabilisticMinInliersStoppingCondition
extends Object
implements RANSAC.StoppingCondition

Stopping condition that tests the number of matches against a percentage threshold of the whole data. If the number exceeds the threshold, then the model is considered to fit.

Constructor Summary

Constructors

Constructor and Description
ProbabilisticMinInliersStoppingCondition(double desiredErrorProbability) Constructor with defaults for bad inlier probability and percentage inliers.
ProbabilisticMinInliersStoppingCondition(double desiredErrorProbability, double inlierIsBadProbability, double percentageInliers) Default constructor.

Method Summary

Modifier and Type	Method and Description
boolean	finalFitCondition(int numInliers) Should the model be considered to fit after the final iteration has passed?
boolean	init(List<?> data, EstimatableModel<?,?> model) Initialise the stopping condition if necessary.
boolean	shouldStopIterations(int numInliers) Should we stop iterating and return the model?

Methods inherited from class java.lang.Object

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

Constructor Detail

ProbabilisticMinInliersStoppingCondition

public ProbabilisticMinInliersStoppingCondition(double desiredErrorProbability,
                                                double inlierIsBadProbability,
                                                double percentageInliers)

Default constructor.

Parameters:

desiredErrorProbability - The desired error rate

inlierIsBadProbability - The probability an inlier is bad

percentageInliers - The percentage of inliers in the data

ProbabilisticMinInliersStoppingCondition

public ProbabilisticMinInliersStoppingCondition(double desiredErrorProbability)

Constructor with defaults for bad inlier probability and percentage inliers.

Parameters:

desiredErrorProbability - The desired error rate

Method Detail

init

public boolean init(List<?> data,
                    EstimatableModel<?,?> model)

Description copied from interface: RANSAC.StoppingCondition

Initialise the stopping condition if necessary. Return false if the initialisation cannot be performed and RANSAC should fail

Specified by:

init in interface RANSAC.StoppingCondition

Parameters:

data - The data being fitted

model - The model to fit

Returns:

true if initialisation is successful, false otherwise.

finalFitCondition

public boolean finalFitCondition(int numInliers)

Description copied from interface: RANSAC.StoppingCondition

Should the model be considered to fit after the final iteration has passed?

Specified by:

finalFitCondition in interface RANSAC.StoppingCondition

Parameters:

numInliers - number of inliers in the final model

Returns:

true if the model fits, false otherwise

shouldStopIterations

public boolean shouldStopIterations(int numInliers)

Description copied from interface: RANSAC.StoppingCondition

Should we stop iterating and return the model?

Specified by:

shouldStopIterations in interface RANSAC.StoppingCondition

Parameters:

numInliers - number of inliers in this iteration

Returns:

true if the model is good and iterations should stop