Prev: Binarisation
Next: Using Matlab Builder JA
From: Khan on 15 Mar 2010 10:13

> Hello Every body
> I am working on a project to search similarities between textures of an image.
> I do not know much about matlab, so please explain the following to me as I am confused about it...
>
> for i = 0:nbdep-1
> a = [max(1,1+i*direction(1)),max(1,1+i*direction(2))];
> b = [min(lign,lign+i*direction(1)),min(colo,colo+i*direction(2))];
> c = [max(1,1-i*direction(1)),max(1,1-i*direction(2))];
> d = [min(lign,lign-i*direction(1)),min(colo,colo-i*direction(2))];
>
> covario(i+1) = distance(imag(a(1):b(1),a(2):b(2)),imag(c(1):d(1),c(2):d(2)),dis);
> end
>
> plot(0:nbdep-1,covario);
> deriv = sign(covario(1:nbdep-1)-covario(2:nbdep));
> signe = deriv(1:nbdep-2)-deriv(2:nbdep-1);
>
> signe = signe<0;
> signe = [signe, 0,0];
> hold on
>
> maxiloco = find((signe.*covario)>0);
> maxiloco = [1, maxiloco];
> plot(maxiloco, covario(maxiloco),'*')
> nbmaxi = size(maxiloco,2);
> moyenne = abs(maxiloco(1:nbmaxi-1)-maxiloco(2:nbmaxi));
> moyenne = mean(moyenne(find(moyenne>5)))
>
> Thanks
 | 
Pages: 1
Prev: Binarisation
Next: Using Matlab Builder JA