Package boofcv.alg.feature.describe.llah

Class LlahHasher

java.lang.Object

boofcv.alg.feature.describe.llah.LlahHasher

Direct Known Subclasses:

LlahHasher.Affine, LlahHasher.CrossRatio

public abstract class LlahHasher
extends Object

Functions related to computing the hash values of a LLAH feature. This is done by looking at the invariant geometry between points and computing a hash function from their values.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	LlahHasher.Affine
static class	LlahHasher.CrossRatio

Field Summary

Fields

Modifier and Type	Field	Description
static int	DEFAULT_HASH_K	Recommended K from the paper
static int	DEFAULT_HASH_SIZE	The recommended hash size from the paper
protected double[]	samples	Defines the look up table.

Constructor Summary

Constructors

Modifier	Constructor	Description
protected	LlahHasher(long hashK, int hashSize)	Configures the hash function.

Method Summary

Modifier and Type	Method	Description
void	computeHash(List<Point2D_F64> points, LlahFeature output)	Computes the hashcode and invariant values.
protected abstract double	computeInvariant(Combinations<Point2D_F64> combinator)	Computes the invariants given the set of points
void	computeInvariants(List<Point2D_F64> points, double[] invariants, int offset)	Stores the computed invariants into an array
int	discretize(double invariant)	Computes the discrete value from the continuous valued invariant
protected abstract int	getInvariantSampleSize()	Number of points required to compute the invariants
int	getNumberOfInvariants(int numPoints)	Returns the number of invariants given the number of points.
int	getNumValues()	Returns the number of possible values
void	learnDiscretization(int[] histogram, int histLength, double histMaxValue, int numDiscrete)	Create a lookup table by sorting then sampling the invariants.

Methods inherited from class java.lang.Object

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

Field Details

DEFAULT_HASH_K

public static int DEFAULT_HASH_K

Recommended K from the paper

DEFAULT_HASH_SIZE

public static int DEFAULT_HASH_SIZE

The recommended hash size from the paper

samples

protected double[] samples

Defines the look up table. A binary search is used to effectively find the index of a value

Constructor Details

LlahHasher

protected LlahHasher(long hashK,
 int hashSize)

Configures the hash function. See JavaDoc for info on variables

Method Details

getNumberOfInvariants

public int getNumberOfInvariants(int numPoints)

Returns the number of invariants given the number of points.

Parameters:

numPoints - Number of points the hash function is computed from

Returns:

Number of invariants the feature will have

computeHash

public void computeHash(List<Point2D_F64> points,
 LlahFeature output)

Computes the hashcode and invariant values. Stores result in output

Parameters:

points - Set of points. Must be ≥ 4.

computeInvariants

public void computeInvariants(List<Point2D_F64> points,
 double[] invariants,
 int offset)

Stores the computed invariants into an array

getInvariantSampleSize

protected abstract int getInvariantSampleSize()

Number of points required to compute the invariants

computeInvariant

protected abstract double computeInvariant(Combinations<Point2D_F64> combinator)

Computes the invariants given the set of points

discretize

public int discretize(double invariant)

Computes the discrete value from the continuous valued invariant

learnDiscretization

public void learnDiscretization(int[] histogram,
 int histLength,
 double histMaxValue,
 int numDiscrete)

Create a lookup table by sorting then sampling the invariants. This will have the desired property of having a denser set of points where there is a higher density of values. A histogram is required instead of raw values because it becomes intractable quickly for even only a few documents if it's an array..

Parameters:

histogram - Histogram of invariant values from 0 to maxValue

histLength - Histogram length.

histMaxValue - The maximum value in the histogram

numDiscrete - Number of possible discrete values. Larger values indicate higher resolution in discretation

getNumValues

public int getNumValues()

Returns the number of possible values