Accumulate

Include <mln/core/algorithm/accumulate.hpp>

1. template<class V, class BinaryFunction>
   V accumulate(InputImage in, V init, BinaryFunction op)
   

2. template<class Accumulator>
   auto accumulate(InputImage in, Accumulator accu)
   

3. template<class Accumulator>
   auto accumulate(InputImage in, Accumulator accu, Extractor ex)
   

   Computes a fold left of the values of the image over an operation. 1) Uses a pair binary function and an initial value? 2) Use an accumulator/extractor pair.

   (1) is equivalent to the following code:

   for (auto&& v : in.values())
       init = op(init, v);
   return init;
   

   (2) is equivalent to the following code:

   for (auto&& v : in.values())
       accu.take(v);
   return accu.to_result();
   

   (3) is equivalent to the following code:

   for (auto&& v : in.values())
       accu.take(v);
   return ex(accu);
   

   Parameters:

   • in – The input image.

   • op – A binary operator

   • init – The initial value to sum

   • accu – The accumulator or feature

   • ex – The value extractor from the accumulator

   Template Parameters:

   • InputImage – A model of InputImage

   • Accumulator – A model of AccumulatorLike

Examples

1. Sum up the values of an image:

   mln::image2d<uint8_t> ima = { {1, 2, 3}, {4, 5, 6} };
   int s = mln::accumulate(ima, 0, std::plus<int>());
   

2. Sum up the values of an image with an accumulator:

   #include <mln/accu/accumulators/sum.hpp>
   
   mln::image2d<uint8_t> ima = { {1, 2, 3}, {4, 5, 6} };
   int s = mln::accumulate(ima, mln::accu::features::sum<>());
   

Complexity

Linear in the number of pixels.