vgfilt − Gaussian filter

vgfilt if=<ifile> of=<ofile> [xs=<sigma in x direction>] [ys=<sigma in y direction>] [zs=<sigma in z direction>] [-mm] [-m] [-v]

vgfilt applies a Gaussian filter to a 3D image. The required input file is the image that needs to be filtered (if=). The output (of=) is the filtered image in the same format as the input image.

The standard deviation of the Gaussian kernel in the x, y and z dimensions can be set by xs=, ys= and zs=, respectively. The default value for these parameters is 2. If the standard deviation for a dimension is set to 0 then the convolution is not performed in that dimension. Therefore, for a 2D image zs= should be set to 0. If the filter parameter for any dimension is less than 0.5 pixels then it is treated as if it were 0; i.e., no filtering.

For DICOM images vgfilt automatically corrects for a different voxel size in z dimension for anisotropic images. Also, using the -mm flag, the standard deviation is defined in millimeters instead of pixels.

Using the -m flag, image border-crossings are evaluated using mirror-images (default is zero-padding). -v is the verbose flag.

if=

input image

of=

output image

xs=

The standard deviation (sigma) in the x dimension

zs=

The standard deviation (sigma) in the z dimension

ys=

The standard deviation (sigma) in the y dimension

-mm

define sigmas in millimeters instead of pixels

-m

use mirror-images to evaluate border-crossings instead of zero-padding

-v

verbose mode

S.J.M. Becx and A. P. Reeves