ivar.m

来自澳大利亚Qeensland大学的计算机视觉Matlab工具箱。 This Toolbox provides a number of functions that are useful in co

%IVAR	Fast neighbourhood variance/kurtosis/skewness 
%
%	V = IVAR(IM, SE, OP)
%	V = IVAR(IM, SE, OP, EDGE)
%
%	Computes the specified statistic over the pixel neighbourhood
%	which becomes the corresponding output pixel value.
%		OP is 'var', 'kurt', 'skew'
%
%	Edge handling flags control what happens when the processing window
%	extends beyond the edge of the image. 	EDGE is either
%		'border' the border value is replicated
%		'none'	 pixels beyond the border are not included in the window
%		'trim'	 output is not computed for pixels whose window crosses
%			 the border, hence output image had reduced dimensions.
%		'wrap'	 the image is assumed to wrap around
%
%	Copyright (c) Peter Corke, 1999  Machine Vision Toolbox for Matlab

%		pic 4/95