imgfiltg 0.4.0 © 2006-2013 by Turku PET Centre
Program for applying 2D or 3D Gaussian filter for a static or dynamic PET
image (ECAT 6.3, 7.x or Analyze 7.5 format).
SD of Gaussian filter must be specified by user, and SD=FWHM/2.355.
2D convolution is slow in case of large image and large filter size,
but more accurate than fast 2D and 3D algorithms.
Command line arguments:
1) dynamic PET image file
2) Gaussian filtered image file
3) SD of Gaussian filter (by default in pixels; see below)
Options:
-method=<<2D>|<fast2D>|<fast3D>>
Specify the Gaussian filtering method; default is the 2D convolution
method which is slow but accurate; fast2D is a less accurate but fast
2D Gaussian convolution algorithm with IIR approximation;
fast3d is 3D Gaussian convolution algorithm with IIR approximation.
-border=<fill|zero>
Image is extrapolated symmetrically before smoothing (fill, default),
or data outside image borders is assumed to be zeroes (zero).
Applies only to 2D method.
-fsize=<filter size>[pxl|mm|cm]
Size of filter kernel, by default 6*SD; size can be be given in pixels
(default), mm, or cm. Final size in pixels will be automatically
changed to the next odd number.
Applies only to 2D method.
-steps=<nr>
More steps provides better accuracy but longer execution time;
default is 4. Applies only to fast2D and fast3D methods.
-pixelsize=<value>
Set image pixel size (mm), if missing or wrong in image data.
-h or --help
Print this message and exit.
--version or --build
Print software build information and exit.
--silent
Program works silently, printing only error and warning messages.
--verbose
Program prints more information about what it is doing.
Example #1a: Gaussian SD in pixels
imgfiltg s5998dy1.v s5998dy1_filt.v 6pxl
Example #1b: Gaussian with SD of 6 millimeters and 40 mm filter size
imgfiltg -fsize=40mm s5998dy1.v s5998dy1_filt.v 6mm
See also:
imgdysmo,
eflexseg,
ecatthrs,
ebkgrm,
fvar4img,
prfpeak,
fitprfs
Keywords: image, gaussian, filter, FWHM
This program comes with ABSOLUTELY NO WARRANTY. This is free software, and
you are welcome to redistribute it under GNU General Public License.
