boofcv.alg.denoise.wavelet

Class DenoiseSureShrink_F32

java.lang.Object

boofcv.alg.denoise.wavelet.SubbandShrink<ImageFloat32>

boofcv.alg.denoise.wavelet.DenoiseSureShrink_F32

All Implemented Interfaces:

DenoiseWavelet<ImageFloat32>

public class DenoiseSureShrink_F32
extends SubbandShrink<ImageFloat32>

SureShrink denoises wavelets using a threshold computed by minimizing Stein's Unbiased Risk Estimate (SURE). In practice a hybrid approach was found to work best where either the Stein threshold or the universal threshold proposed by VisuShrink is used.

This implementation computes a threshold for each subband.

D. Donoho, L. Johnstone, "Adapting to Unknown Smoothness via Wavelet Shrinkage" Journal of the American Statistical Association, Vol. 90, No. 432, December 1995, pp. 1200-1224

Author:

Peter Abeles

Field Summary

Fields inherited from class boofcv.alg.denoise.wavelet.SubbandShrink

rule

Constructor Summary

Constructors

Constructor and Description
DenoiseSureShrink_F32()

Method Summary

Methods

Modifier and Type	Method and Description
protected Number	computeThreshold(ImageFloat32 subband) Compute the threshold for the specified subband.
void	denoise(ImageFloat32 transform, int numLevels) Removes noise from the multi-level wavelet transform.

Methods inherited from class boofcv.alg.denoise.wavelet.SubbandShrink

performShrinkage

Methods inherited from class java.lang.Object

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

Constructor Detail

DenoiseSureShrink_F32

public DenoiseSureShrink_F32()

Method Detail

computeThreshold

protected Number computeThreshold(ImageFloat32 subband)

Description copied from class: SubbandShrink

Compute the threshold for the specified subband.

Specified by:

computeThreshold in class SubbandShrink<ImageFloat32>

Parameters:

subband - Subband whose threshold is being computed.

Returns:

denoise

public void denoise(ImageFloat32 transform,
           int numLevels)

Description copied from interface: DenoiseWavelet

Removes noise from the multi-level wavelet transform.

Parameters:

transform - Transform of the original image.

numLevels - NUmber of levels in the transform.