DeconvolutionLab

How to call DeconvolutionLab from a Javascript script or an ImageJ macro?

We provide below a JavaScript script that illustrates how to run DeconvolutionLab in headless mode. To use the script, download it to your computer and drag-and-drop it onto the ImageJ window.

Depending on the algorithm you want to use, you will need to set a certain number of parameters. We refer to the table below for the full list of parameters.

Example

The script can be downloaded here.

// ***************************************************************************
// DeconvolutionLab
//
// Example of a call from a Javascript macro
// Regularized inverse filter with various epsilon
// ***************************************************************************

IJ.open("/Users/sage/Desktop/Test/input.tif");
IJ.open("/Users/sage/Desktop/Test/psf.tif");

var epsilon =-5;
do {
	IJ.selectWindow("input.tif");
	var chrono = System.currentTimeMillis();
	IJ.run("DeconvolutionLab ", 
		" namealgo='Inverse filtering' epsilon=10e" + epsilon +
		" namepsf='psf.tif' normalizepsf=true recenterpsf=true" +
		" usefftw=false logpriority=1 nameoutput='output'"
	);
	while ((imp=WindowManager.getImage('output')) == null);
	imp.setTitle("output/" + epsilon);
	IJ.log( "Deconvolution Time: " + ((System.currentTimeMillis() - chrono)/1000) + " s." ) ;
	epsilon += 1;	

} while(epsilon < -1);

List of the parameters

Parameter	Type	Description	Default value	Range
Common Parameters
usemacro	Boolean	useMacro	true
namealgo	String	Name of the algorithm	Should be provided
namepsf	String	Name of ImageJ window for the PSF	Should be provided
normalizepsf	Boolean	Normalize PSF	true
recenterpsf	Boolean	Flip PSF quadrants (activate if PSF is centered)	true
monitorperf	Boolean	Computation of the Signal-to-Error ratio	false
nameref	String	Ground-truth image
folderperfmon	String	Folder for performance monitoring
useseed	Boolean	useSeed	false
seed	Float	seed	0
background	Float	Background value	0
subtractbackground	Boolean	Subtract Background	false
estimatebackground	Boolean	estimateBackground	true
savevideo	Boolean	Save the video monitoring	false
foldervideo	String	Folder to save the video monitoring
cropselection	Boolean	Test on a cropped selection	false
usewindowing	Boolean	Usage of the windowing	false
addborders	Boolean	Add borders for windowing	false
percentwindowing	Float	Percent of the windowing	0
percentborder	Float	Percent of the border window	0
windowingfunction	Integer	Windowing function	0	ZEROPADDING=0, HANN=1
processmultifile	Boolean	Process in batch mode	false
folderbatchinput	String	Folder for the batch input
folderbatchoutput	String	Folder for the batch output
usefftw	Boolean	Use FFTW	false
logprioritylevel	Integer	Level of log messages	0	MUTE, SILENT, NORMAL, VERBOSE
nameoutput	String	Name of the output
Convolution
addpoissonnoise	Boolean	Add Poisson noise	false
addgaussiannoise	Boolean	Add Gaussian noise	false
addbackground	Boolean	Add background noise	false
gamma	Float	Gain of the Poisson noise	1	0 ... 65535
sigma	Float	Standard deviation of the Gaussian noise	1	0 ... 10000
mu	Float	Value of the background	0	0 ... 1e6
Inverse Filtering
epsilon	Float	Threshold	1e-6	0 ... 1e6
Regularized Inverse Filtering
regparamest	Boolean	Automatic adjustment of the regularization parameter	true
lambda	Float	Regularization parameter	1	0 ... 1e6
noiseparamest	Boolean	Automatic estimation of the noise parameters	true
gamma	Float	Gain of the Poisson noise	1	1e-6 ... 1e6
sigma	Float	Standard deviation of the Gaussian noise	1	0 ... 1e6
Landweber
initialest	String	Initial estimate
useinputimg	Boolean	Use input image	true
k	Integer	Number of iterations (maximum)	10	1 ... 1e6
tau	Float	Step size	1	0 ... 1e6
autostepsize	Boolean	Automatic step size	true
onego	Boolean	Fixed number of iterations	false
posconstraint	Boolean	Positivity constraint	false
stopcrit	Integer	Early-stopping criterion	0	FIXEDNUMITER=0, RESIDUAL=1, UPDATE=2
sigma	Float	Standard deviation of noise	1	0 ... 1e3
estimatesigma	Boolean	Estimated standard deviation of noise	false
tolerance	Float	Tolerance		0 ... 1
Thresholded Landweber
lambda	Float	Regularization parameter	1	0f ... 1e6
wavfamily	Integer	Wavelet family	0	HAAR=0, DAUBECHIES2=1, DAUBECHIES4=2, DAUBECHIES6=3, FRACT_SPLINE=4
j	Integer	Number of decomposition levels	3	1 ... 10
threshtype	Integer	Thresholding method	0	SOFT=0, HARD=1
randomshifts	Boolean	Usage of the random shifts method	true
stopcrit	Integer	Early-stopping criterion	0	FIXEDNUMITER=0, RESIDUAL=1, UPDATE=2
initialest	String	Initial estimate
useinputimg	Boolean	Use input image	true
k	Integer	Number of iterations (maximum)	10	1 ... 1e6
tau	Float	Step size	1	0 ... 1e6
autostepsize	Boolean	Automatic step size	true
sigma	Float	Standard deviation of noise		0 ... 1e3
estimateSigma	Boolean	Estimated standard deviation of noise	false
tolerance	Float	Tolerance	1e-3	0 ... 1
FISTA
lambda	Float	Regularization parameter	1	0 ... 1e6
wavfamily	Integer	Wavelet family	0	HAAR=0, DAUBECHIES2=1, DAUBECHIES4=2, DAUBECHIES6=3, FRACT_SPLINE=4
J	Integer	Number of decomposition levels	3	1 ... 10
threshtype	Integer	Thresholding method	0	SOFT=0, HARD=1
randomshifts	Boolean	Usage of the random shifts method	true
stopcrit	Integer	Early-stopping criterion	0	FIXEDNUMITER=0, RESIDUAL=1, UPDATE=2
initialest	String	Initial estimate
useinputimg	Boolean	Use input image	true
k	Integer	Number of iterations (maximum)	10	1 ... 1e6
tau	Float	Step size	1	0 ... 1e6
autostepsize	Boolean	Automatic step size	true
sigma	Float	Standard deviation of noise	1	0 ... 1e3
estimatesigma	Boolean	Estimated standard deviation of noise	false
tolerance	Float	Tolerance	1e-3	0 ... 1
Tikhonov Miller
lambda	Float	Regularization parameter	1	0 ...1e6
initialest	String	Initial estimate
useinputimg	Boolean	Use input image	true
k	Integer	Number of iterations (maximum)	10	1 ... 1e6
tau	Float	Step size	1	0 ... 1e6
autostepsize	Boolean	Automatic step size	true
onego	Boolean	Fixed number of iterations	false
posconstraint	Boolean	Positivity constraint	true
stopcrit	Integer	Early-stopping criterion	0	FIXEDNUMITER=0, RESIDUAL=1, UPDATE=2
sigma	Float	Standard deviation of noise	1	0 ... 1e3
estimatesigma	Boolean	Estimated standard deviation of noise	false
tolerance	Float	Tolerance	1e-3	0 ... 1
Richardson Lucy
initialest	String	Initial estimate
useinputimg	Boolean	Use input image	true
k	Integer	Number of iterations (maximum)	10	1 ... 1e6
Richardson Lucy TV
lambda	Float	Regularization parameter	1	0 ... 1e6
initialest	String	Initial estimate
useinputimg	Boolean	Use input image	true
k	Integer	Number of iterations	10	1 ... 1e6
posconstraint	Boolean	Positivity constraint	true