In doing the intensity analysis, our intent was to distinguish Rembrandt from the other two artists by simply using some sort of power spectrum analysis. We knew that most Rembrandt paintings were characterized by dark backgrounds and bright central images. This style can be seen in Rembrandt’s Aristotle contemplating the Bust of Homer.
The viewer’s attention is immediately focused on Aristotle’s face and his concentration on the bust of Homer. The overall darkness of the picture should be easy to distinguish from a typical Monet or Picasso work. Both Monet and Picasso typically painted with vivid colors, making their average intensities very much lighter than Rembrandt’s.
The method we chose to analyze the overall darkness consisted of first plotting a histogram of the intensities and then computing the mean and standard deviation of the histogram.
From these diagrams it is clear that Rembrandt has a lower mean than either Monet or Picasso, and this is the case for all of our test pictures. Thus, we can separate Rembrandt in our algorithm using the mean of the histogram of the intensities of his painting.
After running a few test pictures, it became clear that we needed to set our cut-off parameters for the mean at .25, thus separating Rembrandt from Monet and Picasso. Separating Monet from Picasso, however, turns out to be more challenging and will be left to other tools.