| Histogram |
Mean (histogram’s mean); Variance (histogram’s variance); Skewness (histogram’s skewness); Kurtosis (histogram’s kurtosis); Perc.01% (1% percentile); Perc.10% (10% percentile); Perc.50% (50% percentile); Perc.90% (90% percentile); Perc.99% (99% percentile) |
| Gradient |
GrMean (absolute gradient mean); GrVariance (absolute gradient variance); GrSkewness (absolute gradient skewness); GrKurtosis (absolute gradient kurtosis); GrNonZeros (percentage of pixels with nonzero gradient) |
| Run-length matrix |
RLNonUni (run length nonuniformity); GLevNonU (grey level nonuniformity); LngREmph (long run emphasis); ShrtREmp (short run emphasis); Fraction (fraction of image in runs) |
| Co-occurrence matrix |
AngScMom (angular second moment); Contrast (contrast); Correlat (correlation); SumOfSqs (sum of squares); InvDfMom (inverse difference moment); SumAverg (sum average); SumVarnc (sum variance); SumEntrp (sum entropy); Entropy (entropy); DifVarnc (difference variance); DifEntrp (difference entropy). Features are computed for 5-pixel distance (1, 2, 3, 4, 5) and for 4 various directions (horizontal, 45 degrees, vertical, 135 degrees) |
| Autogressive model |
Teta1 (parameter θ1); Teta2 (parameter θ2); Teta3 (parameter θ3); Teta4 (parameter θ4); Sigma (parameter σ) |
| Wavelet transform |
WavEn (wavelet energy) feature is computed at 5 scales within four frequency bands: low-pass filtering in both directions (LL) assessed the lowest frequencies, low-pass filtering followed by high-pass filtering (LH) assessed horizontal edges, high-pass filtering followed by low-pass filtering (HL) assessed vertical edges; high-pass filtering in both directions (HH) assessed diagonal details. |
