unal.od.jdiffraction.gpu

Class FloatAngularSpectrumGPU

java.lang.Object

unal.od.jdiffraction.gpu.FloatPropagatorGPU

unal.od.jdiffraction.gpu.FloatAngularSpectrumGPU

public class FloatAngularSpectrumGPU
extends FloatPropagatorGPU

Computes wave diffraction through angular spectrum method with single precision.

Since:

JDiffraction 1.2

Author:

Pablo Piedrahita-Quintero (jppiedrahitaq@unal.edu.co), Carlos Trujillo (catrujila@unal.edu.co), Jorge Garcia-Sucerquia (jigarcia@unal.edu.co)

Constructor Summary

Constructors

Constructor and Description
FloatAngularSpectrumGPU(int M, int N, float lambda, float z, float dx, float dy) Creates a new instance of FloatAngularSpectrum.

Method Summary

Modifier and Type	Method and Description
void	diffract(jcuda.driver.CUdeviceptr devField) Performs numerical diffraction of the complex data in devField, leaving the result in devField.
void	diffract(float[] field) Performs numerical diffraction of the complex data in field, leaving the result in field.
float	getDx()
float	getDy()
float	getLambda()
int	getM()
int	getN()
float	getZ()
void	memFree()

Methods inherited from class java.lang.Object

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

Constructor Detail

FloatAngularSpectrumGPU

public FloatAngularSpectrumGPU(int M,
                               int N,
                               float lambda,
                               float z,
                               float dx,
                               float dy)
                        throws java.io.IOException

Creates a new instance of FloatAngularSpectrum. Also performs kernel calculations.

Parameters:

M - Number of data points on x direction.

N - Number of data points on y direction.

lambda - Wavelength.

z - Distance.

dx - Sampling pitch on x direction.

dy - Sampling pitch on y direction.

Throws:

java.io.IOException

Method Detail

diffract

public void diffract(float[] field)

Description copied from class: FloatPropagatorGPU

Performs numerical diffraction of the complex data in field, leaving the result in field. The physical layout of the complex data must be the same as in JTransforms:

field[i * 2 * N + 2 * j] = Re[i][j], field[i * 2 * N + 2 * j + 1] = Im[i][j]; 0 <= i < M, 0 <= j < N

Specified by:

diffract in class FloatPropagatorGPU

Parameters:

field - The complex field to diffract.

diffract

public void diffract(jcuda.driver.CUdeviceptr devField)

Description copied from class: FloatPropagatorGPU

Performs numerical diffraction of the complex data in devField, leaving the result in devField. devField must be a pointer to the data on the GPU memory. The physical layout of the complex data must be the same as in JTransforms:

field[i * 2 * N + 2 * j] = Re[i][j], field[i * 2 * N + 2 * j + 1] = Im[i][j]; 0 <= i < M, 0 <= j < N

Specified by:

diffract in class FloatPropagatorGPU

Parameters:

devField - The complex field to diffract.

memFree

public void memFree()

getM

public int getM()

getN

public int getN()

getZ

public float getZ()

getLambda

public float getLambda()

getDx

public float getDx()

getDy

public float getDy()