vcurvature(1) [debian man page]

vcurvature(1)						      General Commands Manual						     vcurvature(1)

NAME

       vcurvature -

       vcurvature -- 3d curvature computations

SYNOPSIS

Detailed Description
       vcurvature -- 3d curvature computations

       Description
	   3d curvature computations. For each surface voxel, a curvature feature is computed. Surface voxels are defined by a threshold. The
	   algorithm is based on discrete approximations to second-order derivatives and the Weingarten equations.

       Reference:
	   O. Monga, S. Benayoun (1995). 'Using partial derivatives of 3D images to extract typical surface features', Computer Vision and Image
	   Understanding, Vol.61, No.2, pp.171--189.

       Usage

       vcurvature

       Parameters:
	   -in input image
	   -out output image
	   -t iso-value. Default: 0.5
	   -ctype type of output (class | mean | gauss). Default: class
	   -border Whether to process only border voxels (true | false)

       Examples

       Known bugs
	   none.

       Author:
	   G.Lohmann, MPI-CBS, 2004

Author
       Generated automatically by Doxygen for VIA - Volumetric Image Analysis from the source code.

VIA - Volumetric Image Analysis 				  Fri Feb 17 2012						     vcurvature(1)

vdist3d(1Vi)															      vdist3d(1Vi)

NAME

       vdist3d - perform a 3D distance transform

SYNOPSIS

       vdist3d	[-option ...] [infile] [outfile]

DESCRIPTION

       vdist3d	performs a 3D distance transform on the source image to produce a file of a distance transformed destination image.  Each pixel in
       the output image contains the distance to the nearest feature element measured in a Euclidean or Chamfer metric.  If needed,  three  images
       containing the addresses of the nearest foreground pixels will also be produced.

       The input image must be of type VBit. The output image will be of type VFloat. If present, the vector images will be of type VShort.

       Input and output files can be specified on the command line or allowed to default to stdin and stdout.

       If  the	metric is euclidean, and no vectors are needed, the algorithm used is Saito/Toriwaki's algorithm.  If the metric is euclidean, but
       vectors are needed, then Daniellson's algorithm is used.  For the chamfer metric the algorithm by Borgefors is used.   Note  that  Daniell-
       son's  algorithm  is  the  most	expensive  (both  w.r.t  computing  time  as well as w.r.t space).  The second most expensive algorithm is
       Saito/Toriwaki's.  However, it is not possible to obtain the distance vectors from this algorithm.  The cheapest is Borgefors's, which also
       does not yield the distance vectors.

       Note  that  in  some rare cases, Daniellson's algorithm fails.  In that case, use Saito's algorithm (which always works) and do without the
       distance vectors.

LITERATURE

       Toyofumi Saito, Jun-Ichiro Toriwaki, New algorithms for euclidean distance transformation of a  n-dimensional  picture  with  applications,
       Pattern Recognition, Vol. 27, No. 11, pp 1551-1565, 1994.

       Gunilla Borgefors, "Distance Transformations in Arbitrary Dimensions", Computer Vision, Graphics and Image Processing, Vol 27, pp. 321-345,
       1984.

       P.-E. Daniellson, "Euclidean Distance Mapping", Computer Graphics and IMage Processing, Vol. 14, pp. 227-248, 1980.

COMMAND LINE OPTIONS

       vdist3d accepts the following options in any order:

       -help	      Causes vdist3d to print out a message describing its options.

       -metric euclidean/chamfer
		      Determines which metric to use.

       -vector yes/no Determines whether or not a vector image will be produced.  Default: no.

SEE ALSO

       Vista(7Vi)

AUTHOR

       Gabriele Lohmann <lohmann@cns.mpg.de>

Vista Version 2.1						  17 August 1994						      vdist3d(1Vi)