n-dimensional cross-correlation. Generalized version of xcorr2. For 2 dimensional inputs this function is exactly the same as xcorr2, but also works in higher dimensions. Can also be more efficient because it performs convolution using convnFast. Note that xcorr2 is part of the 'Signal Processing Toolbox' and may not be available on all systems. The order of parameters is reversed from normxcorrn. This is to be compatible with the matlab functions normxcorr2 and xcorr2 (which take parameters in different orders). USAGE C = xcorrn( A, T, [shape] ) INPUTS A - first d-dimensional matrix T - second d-dimensional matrix shape - ['full'] 'valid', or 'same' (see convn) OUTPUTS C - correlation matrix EXAMPLE T=gaussSmooth(rand(20),2); A=repmat(T,[3 3]); C1=xcorr2(A,T); C2=xcorrn(A,T); C3=rot90(xcorrn(T,A),2); figure(1); im(C1); figure(2); im(C2); figure(3); im(C3); See also XCORR2, NORMXCORRN, XEUCN, CONVNFAST Piotr's Image&Video Toolbox Version 2.12 Copyright 2008 Piotr Dollar. [pdollar-at-caltech.edu] Please email me if you find bugs, or have suggestions or questions! Licensed under the Lesser GPL [see external/lgpl.txt]