/**
* @file SemanticSegmentation.h
* @brief abstract interface for semantic segmentation algorithms
* @author Erik Rodner
* @date 03/19/2009

*/
#ifndef SEMANTICSEGMENTATIONINCLUDE
#define SEMANTICSEGMENTATIONINCLUDE

#include "vislearning/cbaselib/MultiDataset.h"
#include "vislearning/cbaselib/LocalizationResult.h"
#include "vislearning/cbaselib/CachedExample.h"
#include "vislearning/cbaselib/Example.h"


#define ROADWORKSADD fthrow(NICE::Exception, "addNewExample(const NICE::Vector & newExample, const int & newClassNo): not yet implemented!");
#define ROADWORKSADDNOVEL fthrow(NICE::Exception, "addNovelExamples(): not yet implemented!");
#define ROADWORKSGETNOVEL fthrow(NICE::Exception, "getNovelExamples(): not yet implemented!");


namespace OBJREC
{

/** abstract interface for semantic segmentation algorithms */
class SemanticSegmentation
{

  protected:
    /** accessible class names and information about
        number of classes etc. */
    const ClassNames *classNames;

    /** enum type for imagetype */
    enum
    {
      IMAGETYPE_RGB = 0,
      IMAGETYPE_GRAY
    };

    /** whether to load images with color information */
    int imagetype;
    
    int iterationCountSuffix;

  public:

    /** simple constructor
        @param conf global settings
        @param classNames this ClassNames object while be stored as a attribute
    */
    SemanticSegmentation ( const NICE::Config *conf,
                           const ClassNames *classNames );

    /** simple destructor */
    virtual ~SemanticSegmentation();

    /** this function has to be overloaded by all subclasses
        @param ce image data
        @param segresult result of the semantic segmentation with a label for each
        pixel
        @param probabilities multi-channel image with one channel for each class and
               corresponding probabilities for each pixel
    */
    virtual void semanticseg ( OBJREC::CachedExample *ce,
                               NICE::Image & segresult,
                               NICE::MultiChannelImageT<double> & probabilities ) = 0;

    /**
     * convert different datatypes
     */
    void convertVVectorToExamples ( NICE::VVector &feats,OBJREC::Examples &examples, std::vector<int> &label );
    void convertExamplesToVVector ( NICE::VVector &feats,OBJREC::Examples &examples, std::vector<int> &label );
    void convertExamplesToLSet ( OBJREC::Examples &examples, OBJREC::LabeledSetVector &lvec );
    void convertLSetToExamples ( OBJREC::Examples &examples, OBJREC::LabeledSetVector &lvec, const bool & removeOldDataPointer=false );
    void convertLSetToSparseExamples ( OBJREC::Examples &examples, OBJREC::LabeledSetVector &lvec );


    /** load img from file call localize(CachedExample *ce) etc. */
    void semanticseg ( const std::string & filename,
                       NICE::Image & segresult,
                       NICE::MultiChannelImageT<double> & probabilities );
    
    virtual void addNewExample(const NICE::Vector & newExample, const int & newClassNo)
    {
      ROADWORKSADD;      
    };
    /**
     * @brief Add those examples, which belong to the most novel region seen so far
     *
     * @return void
     **/    
    virtual void addNovelExamples()
    {
      ROADWORKSADDNOVEL;      
    };    
    
    /**
     * @brief Get a pointer to the examples extracted from the most novel region seen so far
     *
     * @return Examples *
     **/        
    virtual const Examples * getNovelExamples() const
    {
      ROADWORKSGETNOVEL;      
    };        
    
    void setIterationCountSuffix( const int & _iterationCountSuffix) { iterationCountSuffix = _iterationCountSuffix; };
    
    

};


} // namespace

#endif