public Dictionary <int, float> getCRHistogram(Bitmap img, bool isCenter, float centering)
{
CenteringRefinement cr = new CenteringRefinement(img, centering);
int imgDimensions = cr.getDimensions(isCenter);
return(quantizeColors(convertToLuv(cr.getRgbHistogram(isCenter)), imgDimensions));
}
public List <String> returnRelevantImages(Bitmap image, bool center)
{
Bitmap img = new Bitmap(image);
//bool center = true;
CenteringRefinement cr = new CenteringRefinement(img, centering);
imgDimensions = cr.getDimensions(center);
quantizedHistogram = ch.getCRHistogram(img, center, centering);
foreach (int x in quantizedHistogram.Keys)
{
normalizedHistogram.Add(x, quantizedHistogram[x] / imgDimensions);
}
for (int i = 0; i < 159; i++)
{
normalizedHistogram.TryGetValue(i, out hist1[i]);
}
foreach (String s in imagePaths)
{
paths.Add(s);
}
// Loop for currentImg
foreach (String s in paths)
{
Dictionary <int, float> currQuantizedHistogram = new Dictionary <int, float>();
Dictionary <int, float> currNormalizedHistogram = new Dictionary <int, float>();
Bitmap currImg = new Bitmap(s);
// CH with Centering Refinement
CenteringRefinement cr2 = new CenteringRefinement(currImg, centering);
//bool cent = false;
currImgDimensions = cr2.getDimensions(center);
currQuantizedHistogram = ch.getCRHistogram(img, center, centering);
sim = 0;
foreach (int x in currQuantizedHistogram.Keys)
{
currNormalizedHistogram.Add(x, currQuantizedHistogram[x] / currImgDimensions);
}
for (int i = 0; i < 159; i++)
{
currNormalizedHistogram.TryGetValue(i, out hist2[i]);
}
sim = ch.computeSimilarity(hist1, hist2, threshold);
if (sim > simThreshold)
{
similarImagesPaths.Add(s);
}
//Debug.WriteLine(sim);
if (center)
{
centeredSim.Add(sim);
}
else
{
nonCenteredSim.Add(sim);
}
}
return(similarImagesPaths);
}
