private static Dictionary <Color, ColorInformation> GetColorPallete(Bitmap bmp)
{
//this is
Dictionary <Color, ColorInformation> pallete = new Dictionary <Color, ColorInformation>();
for (int column = 0; column < bmp.Width; column++)
{
for (int row = 0; row < bmp.Height; row++)
{
Color pixelColor = bmp.GetPixel(column, row);
//this create a color that is in the middle of the color band
Color normalColor = NormalizeColor(pixelColor);
ColorInformation ci = null;
if (pallete.TryGetValue(normalColor, out ci))
{
ci.numberOfPixels++;
}
else
{
ci = new ColorInformation();
ci.numberOfPixels = 1;
ci.PalleteColor = normalColor;
pallete.Add(normalColor, ci);
}
}
}
return(pallete);
}
public static Dictionary <Color, ColorInformation> AnalyzeBitmaps(WriteableBitmap imageWithObject, WriteableBitmap imageWithoutObject)
{
Bitmap bitmapWithObject = BitmapImage2Bitmap(imageWithObject);
Bitmap bitmapWithoutObject = BitmapImage2Bitmap(imageWithoutObject);
//get the pallet for each, these are already normalize into color bands
Dictionary <Color, ColorInformation> backgroundPallete = GetColorPallete(bitmapWithoutObject);
Dictionary <Color, ColorInformation> objectPallete = GetColorPallete(bitmapWithObject);
Dictionary <Color, ColorInformation> resultPallete = new Dictionary <Color, ColorInformation>();
//if the color is absent from the background or 10 times more common in the learning
//bitmp, then it is in the color pallete of the object to learn
foreach (Color color in objectPallete.Keys)
{
ColorInformation backgroundColorInfo = null;
ColorInformation objectColorInfo = null;
backgroundPallete.TryGetValue(color, out backgroundColorInfo);
objectPallete.TryGetValue(color, out objectColorInfo);
if ((backgroundColorInfo == null) || ((backgroundColorInfo.numberOfPixels * 10) < objectColorInfo.numberOfPixels))
{
resultPallete.Add(color, objectColorInfo);
}
}
return(resultPallete);
}
