--- /dev/null
+## Copyright (C) 2004 Josep Mones i Teixidor
+##
+## This program is free software; you can redistribute it and/or modify
+## it under the terms of the GNU General Public License as published by
+## the Free Software Foundation; either version 2 of the License, or
+## (at your option) any later version.
+##
+## This program is distributed in the hope that it will be useful,
+## but WITHOUT ANY WARRANTY; without even the implied warranty of
+## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+## GNU General Public License for more details.
+##
+## You should have received a copy of the GNU General Public License
+## along with this program; If not, see <http://www.gnu.org/licenses/>.
+
+## -*- texinfo -*-
+## @deftypefn {Function File} {@var{B} = } im2col (@var{A}, [@var{m},@var{n}], @var{block_type})
+## @deftypefnx {Function File} {@var{B} = } im2col (@var{A}, [@var{m},@var{n}])
+## @deftypefnx {Function File} {@var{B} = } im2col (@var{A}, 'indexed', ...)
+## Rearranges image blocks into columns.
+##
+## @code{B=im2col(A, [m, n], blocktype)} rearranges blocks in @var{A}
+## into columns in a way that's determined by @var{block_type}, which
+## can take the following values:
+##
+## @table @code
+## @item distinct
+## Rearranges each distinct @var{m}-by-@var{n} block in image @var{A}
+## into a column of @var{B}. Blocks are scanned from left to right and
+## the up to bottom in @var{A}, and columns are added to @var{B} from
+## left to right. If @var{A}'s size is not multiple @var{m}-by-@var{n}
+## it is padded.
+## @item sliding
+## Rearranges any @var{m}-by-@var{n} sliding block of @var{A} in a
+## column of @var{B}, without any padding, so only sliding blocks which
+## can be built using a full @var{m}-by-@var{n} neighbourhood are taken.
+## In consequence, @var{B} has @var{m}*@var{n} rows and
+## (@var{mm}-@var{m}+1)*(@var{nn}-@var{n}+1) columns (where @var{mm} and
+## @var{nn} are the size of @var{A}).
+##
+## This case is thought to be used applying operations on columns of
+## @var{B} (for instance using sum(:)), so that result is a
+## 1-by-(@var{mm}-@var{m}+1)*(@var{nn}-@var{n}+1) vector, that is what
+## the complementary function @code{col2im} expects.
+## @end table
+##
+## @code{B=im2col(A,[m,n])} takes @code{distinct} as a default value for
+## @var{block_type}.
+##
+## @code{B=im2col(A,'indexed',...)} will treat @var{A} as an indexed
+## image, so it will pad using 1 if @var{A} is double. All other cases
+## (incluing indexed matrices with uint8 and uint16 types and
+## non-indexed images) will use 0 as padding value.
+##
+## Any padding needed in 'distinct' processing will be added at right
+## and bottom edges of the image.
+##
+## @seealso{col2im}
+## @end deftypefn
+
+## Author: Josep Mones i Teixidor <jmones@puntbarra.com>
+
+function B = im2col(A, varargin)
+ if(nargin<2 || nargin>4)
+ usage("B=im2col(B [, 'indexed'], [m,n] [, block_type])");
+ endif
+
+ ## check 'indexed' presence
+ indexed=false;
+ p=1;
+ if(ischar(varargin{1}) && strcmp(varargin{1}, "indexed"))
+ if(nargin<3)
+ usage("B=im2col(B [, 'indexed'], [m,n] [, block_type])");
+ endif
+ indexed=true;
+ p+=1;
+ if(isa(A,"uint8") || isa(A,"uint16"))
+ padval=0;
+ else
+ padval=1;
+ endif
+ else
+ padval=0;
+ endif
+
+ ## check [m,n]
+ if(!isvector(varargin{p}))
+ error("im2col: expected [m,n] but param is not a vector.");
+ endif
+ if(length(varargin{p})!=2)
+ error("im2col: expected [m,n] but param has wrong length.");
+ endif
+ m=varargin{p}(1);
+ n=varargin{p}(2);
+ p+=1;
+
+ block_type='sliding';
+ if(nargin>p)
+ ## we have block_type param
+ if(!ischar(varargin{p}))
+ error("im2col: invalid parameter block_type.");
+ endif
+ block_type=varargin{p};
+ p+=1;
+ endif
+
+ ## if we didn't have 'indexed' but had 4 parameters there's an error
+ if(nargin>p)
+ usage("B=im2col(B [, 'indexed'], [m,n] [, block_type])");
+ endif
+
+
+ ## common checks
+ if(!ismatrix(A))
+ error("im2col: A should be a matrix (or vector).");
+ endif
+
+ switch(block_type)
+ case('distinct')
+ ## calc needed padding
+ sp=mod(-size(A)',[m;n]);
+
+ if(any(sp))
+ A=padarray(A,sp,padval,'post');
+ endif
+
+ ## iterate through all blocks
+ B=[];
+ for i=1:m:size(A,1) ## up to bottom
+ for j=1:n:size(A,2) ## left to right
+ ## TODO: check if we can horzcat([],uint8([10;11])) in a
+ ## future Octave version > 2.1.58
+ if(isempty(B))
+ B=A(i:i+m-1,j:j+n-1)(:);
+ else
+ B=horzcat(B, A(i:i+m-1,j:j+n-1)(:));
+ endif
+ endfor
+ endfor
+
+ case('sliding')
+ if(indexed)
+ disp("WARNING: 'indexed' has no sense when using sliding.");
+ endif
+ if(m>size(A,1) || n>size(A,2))
+ error("im2col: block size can't be greater than image size in sliding");
+ endif
+ ## TODO: check if matlab uses top-down and left-right order
+ B=[];
+ for j=1:1:size(A,2)-n+1 ## left to right
+ for i=1:1:size(A,1)-m+1 ## up to bottom
+ ## TODO: check if we can horzcat([],uint8([10;11])) in a
+ ## future Octave version > 2.1.58
+ if(isempty(B))
+ B=A(i:i+m-1,j:j+n-1)(:);
+ else
+ B=horzcat(B, A(i:i+m-1,j:j+n-1)(:));
+ endif
+ endfor
+ endfor
+
+ otherwise
+ error("im2col: invalid block_type.");
+ endswitch
+
+endfunction
+
+%!demo
+%! A=[1:10;11:20;21:30;31:40]
+%! B=im2col(A,[2,5],'distinct')
+%! C=col2im(B,[2,5],[4,10],'distinct')
+%! # Divide A using distinct blocks and reverse operation
+
+%!shared B, A, Bs, As, Ap, Bp0, Bp1
+%! v=[1:10]';
+%! r=reshape(v,2,5);
+%! B=[v, v+10, v+20, v+30, v+40, v+50];
+%! A=[r, r+10; r+20, r+30; r+40, r+50];
+%! As=[1,2,3,4,5;6,7,8,9,10;11,12,13,14,15];
+%! b1=As(1:2,1:4)(:);
+%! b2=As(2:3,1:4)(:);
+%! b3=As(1:2,2:5)(:);
+%! b4=As(2:3,2:5)(:);
+%! Bs=[b1,b2,b3,b4];
+%! Ap=A(:,1:9);
+%! Bp1=Bp0=B;
+%! Bp0([9:10],[2,4,6])=0;
+%! Bp1([9:10],[2,4,6])=1;
+
+%!# bad block_type
+%!error(im2col(A,[2,5],'wrong_block_type'));
+
+%!# distinct
+%!assert(im2col(A,[2,5],'distinct'), B);
+
+%!# padding
+%!assert(im2col(Ap,[2,5],'distinct'), Bp0);
+%!assert(im2col(Ap,'indexed',[2,5],'distinct'), Bp1);
+
+%!# now sliding
+%!assert(im2col(As,[2,4]), Bs);
+%!assert(im2col(As,[2,4],'sliding'), Bs);
+%!assert(im2col(As,[3,5],'sliding'), As(:));
+
+%!# disctint uint8 & uint16
+%!assert(im2col(uint8(A),[2,5],'distinct'), uint8(B));
+%!assert(im2col(uint16(A),[2,5],'distinct'), uint16(B));
+
+%!# padding uint8 & uint16 (to 0 even in indexed case)
+%!assert(im2col(uint8(Ap),[2,5],'distinct'), uint8(Bp0));
+%!assert(im2col(uint8(Ap),'indexed',[2,5],'distinct'), uint8(Bp0));
+%!assert(im2col(uint16(Ap),[2,5],'distinct'), uint16(Bp0));
+%!assert(im2col(uint16(Ap),'indexed',[2,5],'distinct'), uint16(Bp0));
+
+%!# now sliding uint8 & uint16
+%!assert(im2col(uint8(As),[2,4],'sliding'), uint8(Bs));
+%!assert(im2col(uint16(As),[2,4],'sliding'), uint16(Bs));
+
+
+
+
+%
+% $Log$
+% Revision 1.4 2007/03/23 16:14:36 adb014
+% Update the FSF address
+%
+% Revision 1.3 2007/01/04 23:47:43 hauberg
+% Put seealso before end deftypefn
+%
+% Revision 1.2 2007/01/04 23:37:54 hauberg
+% Minor changes in help text
+%
+% Revision 1.1 2006/08/20 12:59:33 hauberg
+% Changed the structure to match the package system
+%
+% Revision 1.3 2005/09/08 02:00:17 pkienzle
+% [for Bill Denney] isstr -> ischar
+%
+% Revision 1.2 2004/09/03 17:37:08 jmones
+% Added support for int* and uint* types
+%
+% Revision 1.1 2004/08/18 14:39:07 jmones
+% im2col and col2im added
+%
+%
git://git.creatis.insa-lyon.fr / CreaPhase.git / blobdiff