VDOGVisionX V4 NAME

vdog - Difference of Gaussians filter

SYNOPSIS

vdog if= [of=] [r=] [s=] [-2d] [-3d] [-a] [-f] [-mm] [-v] [-d]

DESCRIPTION

vdog is a reference implemntation of the Difference of Gaussiaan (DoG) filter. The s= parameter specifies the sigma for the small Gaussian and the r= parameter specifies the the ratio of small to large Gaussians (default 1.6).

The filter is implemented for both 2D and 3D images which may be explicitly set with the -2d and -3d options. If neither option is set then 2D filtering is performed on images without frame markers or having only one frame and 3D mode is set multiple images with frame markers.

r=1.6 approxiation for Laplacian of Gaussian, r=5 approximation for human visual system

CONSTRAINTS

Multiple channel images not supported.

OPTIONS

	if=		input image file
	of=		ouput image file
	s=		sigma of the smaller Gaussian function (default 1.5)
	r=		ratio of large to small sigmas (default 1.6).
	-2d		2D mode, only filter the first two image dimensions
	-3d		3D mode, only filter in all three image diemensions
	-a		autoscale the output to a byte image
	-f		keep the output image in float format
	-v		verbose mode
	-mm		sigma is expressed in mm rather than pixels (for DICOM images)
	-d		debug mode

AUTHOR

A. P. Reeves