There are different m files for different filter to use. perwit.m:=uses window Wx=[1 0 -1 1 0 -1]/3 and Wy=Wx and then convolves this window with the image. sobel.m:=uses window W=[1 0 -1 2 0 -2 1 0 -1]/4 lablacian.m=uses window W=[0 1 0 1

CCA calculate canonical correlations [Wx Wy r] = cca(X,Y) where Wx and Wy contains the canonical correlation vectors as columns and r is a vector with corresponding canonical correlations. The correlations are sorted in descending order. X a