[Insight-users] Design question on compare histograms

[Martin Kavec]

Hi all,

I am working on a histogram comparison class, which would have the
same functionality as in OpenCV: take two histograms and return a
metric of their similarity. I am facing however a principal design
problems so I would like to ask for help here:

1. what would be the best class to inherit from?
2. I am templating the class by input image type, which in this case
has to be the same for both reference and source
3. I would also like to provide a methods to NormalizeHeight and
NormalizeRange of the histograms

Particularly with the Normalize methods I had so far problems, where I
could not find a way through. I need an iterator to go through the
histogram, but

	HistogramType::Iterator itr_reference = m_ReferenceHistogram->Begin();

gives me

itk::Statistics::Histogram<TMeasurement>::Iterator::Iterator' : cannot
access protected member declared in class
itk::Statistics::Histogram<TMeasurement>::Iterator'

Thanks for suggestions,

Martin