boofcv.alg.feature.describe.brief

Class FactoryBriefDefinition

java.lang.Object

boofcv.alg.feature.describe.brief.FactoryBriefDefinition

public class FactoryBriefDefinition
extends java.lang.Object

Creates different brief descriptors. The original paper proposed several different configurations. The best one were randomly generated with a gaussian distribution.

Constructor Summary

Constructors

Constructor and Description
FactoryBriefDefinition()

Method Summary

Modifier and Type	Method and Description
static BinaryCompareDefinition_I32	gaussian(java.util.Random rand, int radius, int numPairs) Creates a descriptor by randomly selecting points inside a square region using a Gaussian distribution with a sigma of (5/2)*radius.
static BinaryCompareDefinition_I32	gaussian2(java.util.Random rand, int radius, int numPairs)

Methods inherited from class java.lang.Object

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

Constructor Detail

FactoryBriefDefinition

public FactoryBriefDefinition()

Method Detail

gaussian

public static BinaryCompareDefinition_I32 gaussian(java.util.Random rand,
                                                   int radius,
                                                   int numPairs)

Creates a descriptor by randomly selecting points inside a square region using a Gaussian distribution with a sigma of (5/2)*radius. This is done exactly as is described in the paper where twice as many points are sampled as are compared..

Parameters:

rand - Random number generator.

radius - Radius of the square region. width = 2*radius+1.

numPairs - Number of sample point pairs.

Returns:

Definition of a BRIEF feature.

gaussian2

public static BinaryCompareDefinition_I32 gaussian2(java.util.Random rand,
                                                    int radius,
                                                    int numPairs)