tvdenoise.c File Reference

Total variation regularized denoising demo for IPOL. More...

#include <math.h>
#include <stdio.h>
#include <string.h>
#include <ctype.h>
#include "num.h"
#include "tvreg.h"
#include "imageio.h"

Go to the source code of this file.

Data Structures
struct	image
	struct representing an image More...
struct	programparams
	struct of program parameters More...

Macros
#define	DISPLAY_SCALING   255
	Display intensities in the range [0,DISPLAY_SCALING].
#define	LAMBDA_TUNE_ITERATIONS   5
	Number of iterations for tuning lambda.
#define	JPEGQUALITY   95
	Quality for writing JPEG images.
#define	IMAGEIO_NUM   (IMAGEIO_DOUBLE)

Functions
void	PrintHelpMessage ()
	Print program explanation and usage.
int	Denoise (image u, image f, const char *Model, num Sigma, num Lambda)
	TV regularized denoising. More...
num	ComputeRmse (image f, image u)
	Compute the root mean-square-error between two images.
int	IsGrayscale (image f)
	Test whether image is grayscale.
static int	ParseParams (programparams *Param, int argc, char *argv[])
int	main (int argc, char **argv)
int	LambdaTune (tvregopt *Opt, image u, image f, const char *Model, num Sigma)
	Tune lambda according to the discrepancy principle. More...

Detailed Description

Total variation regularized denoising demo for IPOL.

Author

Pascal Getreuer getre.nosp@m.uer@.nosp@m.gmail.nosp@m..com

Copyright (c) 2011-2012, Pascal Getreuer All rights reserved.

This program is free software: you can use, modify and/or redistribute it under the terms of the simplified BSD License. You should have received a copy of this license along this program. If not, see http://www.opensource.org/licenses/bsd-license.html.

Definition in file tvdenoise.c.

Function Documentation

int Denoise	(	image	u,
		image	f,
		const char *	Model,
		num	Sigma,
		num	Lambda
	)

TV regularized denoising.

Parameters

u	denoised image
f	given noisy image
Model	string specifying either "gaussian", "laplace", or "poisson"
Sigma	the standard deviation of the noise
Lambda	the fidelity stength

Returns

1 on success, 0 on failure

Definition at line 237 of file tvdenoise.c.

int LambdaTune	(	tvregopt *	Opt,
		image	u,
		image	f,
		const char *	Model,
		num	Sigma
	)

Tune lambda according to the discrepancy principle.

Parameters

Opt	tvreopt options variable
u	the denoised image
f	the given noisy image
Model	string specifying either "gaussian", "laplace", or "poisson"
Sigma	the standard deviation of the noise

Returns

1 on success, 0 on failure

This routine tunes lambda according to the discrepancy principle. Empirical estimates of the optimal lambda value are used as the initialization.

The TV denoising computation itself is performed by TvRestore().

Definition at line 171 of file tvdenoise.c.