TABLE 3

Relevant GLCM Texture Features, Originally Described by Haralick et al. (10)

Feature nameFormula
ContrastEmbedded Image
Sum of squares: varianceEmbedded Image
Sum varianceEmbedded Image
Difference entropyEmbedded Image
  • Embedded Imageθ) corresponds to entry in GLCM, designating probability of pixel with gray-level of i being distance d pixels away from neighboring pixel with gray-level of j in θ direction. In this study, we fixed d at 1 pixel and averaged over available directions for θ. Ng represents number of gray levels. x and y are row and column indices of GLCM, and Embedded Imageis probability of 2 indices summing to Embedded Image, whereas Embedded Imageis probability that difference 2 indices is Embedded Image; μ is mean of p.