The image histogram, a first-order statistical measure of the distribution of image pixel amplitudes, is a widely used source of histogram-based features. These play an important role in region-based image segmentation. The shape of an image histogram provides many clues as to the character of the sample values. Commonly used histogram-based features include the mode (i.e., the most frequently occurring amplitude value), variance, entropy, energy, kurtosis, and skewness [
Har73]. This calculates the local histograms.