convolve.hh

// Copyright (C) 2007, 2008, 2009 EPITA Research and Development Laboratory (LRDE)
//
// This file is part of Olena.
//
// Olena 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, version 2 of the License.
//
// Olena 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 Olena.  If not, see <http://www.gnu.org/licenses/>.
//
// As a special exception, you may use this file as part of a free
// software project without restriction.  Specifically, if other files
// instantiate templates or use macros or inline functions from this
// file, or you compile this file and link it with other files to produce
// an executable, this file does not by itself cause the resulting
// executable to be covered by the GNU General Public License.  This
// exception does not however invalidate any other reasons why the
// executable file might be covered by the GNU General Public License.

#ifndef MLN_LINEAR_CONVOLVE_HH
# define MLN_LINEAR_CONVOLVE_HH


# include <mln/core/concept/image.hh>
# include <mln/core/concept/weighted_window.hh>
# include <mln/linear/ch_convolve.hh>
# include <mln/accu/convolve.hh>
# include <mln/extension/adjust_duplicate.hh>


namespace mln
{

  namespace linear
  {

    template <typename I, typename W>
    mln_ch_convolve(I, W)
    convolve(const Image<I>& input, const Weighted_Window<W>& w_win);



# ifndef MLN_INCLUDE_ONLY

    // Tests.

    namespace internal
    {

      template <typename I, typename W>
      void
      convolve_tests(const Image<I>& input,
                     const Weighted_Window<W>& w_win)
      {
        mln_precondition(exact(input).is_valid());
        mln_precondition(exact(w_win).is_valid());
        (void) input;
        (void) w_win;
      }

    }  // end of namespace mln::linear::internal


    // Implementation.

    namespace impl
    {

      namespace generic
      {

        template <typename I, typename W>
        mln_ch_convolve(I, W)
        convolve(const Image<I>& input_,
                 const Weighted_Window<W>& w_win_)
        {
          trace::entering("linear::impl::generic::convolve");

          const I& input = exact(input_);
          const W& w_win = exact(w_win_);
          internal::convolve_tests(input, w_win);

          extension::adjust_duplicate(input, w_win);
          
          typedef mln_ch_convolve(I, W) O;
          O output;
          initialize(output, input);

          accu::convolve<mln_value(I), mln_weight(W)> a;

          mln_piter(I) p(input.domain());
          mln_qiter(W) q(w_win, p);

          for_all(p)
          {
            a.init();
            for_all(q) if (input.has(q))
              a.take(input(q), q.w());
            output(p) = a.to_result();
          }
          
          trace::exiting("linear::impl::generic::convolve");
          return output;
        }

      } // end of namespace mln::linear::impl::generic


      template <typename I, typename W>
      mln_ch_convolve(I, W)
      convolve_fastest(const Image<I>& input_,
                       const Weighted_Window<W>& w_win_)
      {
        trace::entering("linear::impl::convolve_fastest");

        const I& input = exact(input_);
        const W& w_win = exact(w_win_);
        internal::convolve_tests(input, w_win);

        extension::adjust_duplicate(input, w_win);

        typedef mln_ch_convolve(I, W) O;
        O output;
        initialize(output, input);
        mln_pixter(O) p_out(output);

        accu::convolve<mln_value(I), mln_weight(W)> a;

        mln_pixter(const I)    p(input);
        mln_qixter(const I, W) q(p, w_win);

        for_all_2(p, p_out)
          {
            a.init();
            unsigned i = 0;
            for_all(q)
              a.take(q.val(), w_win.w(i++));
            p_out.val() = a.to_result();
          }

        trace::exiting("linear::impl::convolve_fastest");
        return output;
      }

    } // end of namespace mln::linear::impl


    // Dispatch.

    namespace internal
    {

      template <typename I, typename W>
      mln_ch_convolve(I, W)
      convolve_dispatch(trait::image::speed::any,
                        const Image<I>& input,
                        const Weighted_Window<W>& w_win)
      {
        return impl::generic::convolve(input, w_win);
      }

      template <typename I, typename W>
      mln_ch_convolve(I, W)
      convolve_dispatch(trait::image::speed::fastest,
                        const Image<I>& input,
                        const Weighted_Window<W>& w_win)
      {
        return impl::convolve_fastest(input, w_win);
      }

      template <typename I, typename W>
      mln_ch_convolve(I, W)
      convolve_dispatch(const Image<I>& input,
                        const Weighted_Window<W>& w_win)
      {
        return convolve_dispatch(mln_trait_image_speed(I)(),
                                 input, w_win);
      }

    }  // end of namespace mln::linear::internal


    // Facade.

    template <typename I, typename W>
    mln_ch_convolve(I, W)
    convolve(const Image<I>& input, const Weighted_Window<W>& w_win)
    {
      trace::entering("linear::convolve");

      internal::convolve_tests(input, w_win);

      mln_ch_convolve(I, W) output;
      output = internal::convolve_dispatch(mln_trait_image_speed(I)(),
                                           input, w_win);

      trace::exiting("linear::convolve");
      return output;
    }

# endif // ! MLN_INCLUDE_ONLY

  } // end of namespace mln::linear

} // end of namespace mln


#endif // ! MLN_LINEAR_CONVOLVE_HH