boofcv.alg.interpolate.array

Class Interpolate1D_F32

java.lang.Object

boofcv.alg.interpolate.array.Interpolate1D_F32

Direct Known Subclasses:

PolynomialNeville_F32

public abstract class Interpolate1D_F32
extends Object

Provides much of the basic house keeping needed for interpolating 1D data. Interpolation is done using sample points y[i] = f(x[i]) where x is a monotonically increasing or decreasing function.

Author:

Peter Abeles

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	ascend
protected int	index0
protected int	M
protected int	size
protected float[]	x
protected float[]	y

Constructor Summary

Constructors

Constructor and Description
Interpolate1D_F32(int degree)
Interpolate1D_F32(int degree, float[] x, float[] y, int size)

Method Summary

Methods

Modifier and Type	Method and Description
protected void	bisectionSearch(float val, int lowerLimit, int upperLimit) Searches the x array by bisecting it.
void	changeDegree(int degree) Changes the number of points used in the interpolation.
protected abstract float	compute(float testX) This is where the specific implementation of the interpolation is done.
protected void	hunt(float val) To speed up finding the appropriate indexes to use in the interpolation it can use its previous results to search a smaller region than it would otherwise.
float	process(float testX) Performs interpolation at the sample point.
float	process(int index0, float testX) Performs an interpolation using sample data starting at index0.
void	setInput(float[] x, float[] y, int size) Sets the data that is being interpolated.

Methods inherited from class java.lang.Object

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

Field Detail

size

protected int size

x

protected float[] x

y

protected float[] y

M

protected int M

index0

protected int index0

ascend

protected boolean ascend

Constructor Detail

Interpolate1D_F32

public Interpolate1D_F32(int degree)

Parameters:

degree - The number of points used in the interpolation minus one

Interpolate1D_F32

public Interpolate1D_F32(int degree,
                 float[] x,
                 float[] y,
                 int size)

Parameters:

degree - The number of points used in the interpolation minus one

x - Where the points are sample at. Not modifed. Reference saved.

y - The value at the sample points. Not modifed. Reference saved.

size - The number of points used.

Method Detail

setInput

public void setInput(float[] x,
            float[] y,
            int size)

Sets the data that is being interpolated.

Parameters:

x - Where the points are sample at. Not modifed. Reference saved.

y - The value at the sample points. Not modifed. Reference saved.

size - The number of points used.

process

public float process(float testX)

Performs interpolation at the sample point.

Parameters:

testX - Where the interpolated value is done at.

Returns:

The interpolated value at sampleX.

process

public float process(int index0,
            float testX)

Performs an interpolation using sample data starting at index0. Little checking is done and it is assumed the user knows what he is doing. Interpolation is done using points from index0 to index0 + M - 1

Parameters:

index0 - first sample point used in the interpolation.

testX - Where the interpolated value is done at.

Returns:

The interpolated value at sampleX.

compute

protected abstract float compute(float testX)

This is where the specific implementation of the interpolation is done. It should use points index0 to index0 + M - 1 in its interpolation

Parameters:

testX - Where the interpolated value is done at.

Returns:

The interpolated value at sampleX.

changeDegree

public void changeDegree(int degree)

Changes the number of points used in the interpolation.

Parameters:

degree - Number of points used minus one.

hunt

protected void hunt(float val)

To speed up finding the appropriate indexes to use in the interpolation it can use its previous results to search a smaller region than it would otherwise.

Parameters:

val - The value that is to be interpolated.

bisectionSearch

protected void bisectionSearch(float val,
                   int lowerLimit,
                   int upperLimit)

Searches the x array by bisecting it. This takes advantage of the data being monotonic. This finds a center index which has the following property: x[center] <= val < x[center+1] From that it selects index0 which is center - M/2.

Parameters:

val - The value that is to be interpolated.

lowerLimit - Lower limit for x index.

upperLimit - The largest possible index of x