Table 1 Representative gray-level histogram and gray-level co-occurrence matrix texture features

 Mean

Mean gray-level value

Mean= \( \sum k\frac{k^{\ast }g(k)}{\sum k{g}^k} \)

 Minimum

Minimum gray-level value

Min= Min(k)

 Maximum

Maximum gray-level value

Max= Max(k)

 Skewness

Measure of histogram symmetry

Skew= σ⁻⁴∑_k(k − μ)⁴ ∗ g(k) − 3

 Kurtosis

Measure of histogram flatness

Kurt= σ⁻³∑_k(k − μ)⁴ ∗ g(k)

 Angular Second Moment (ASM)

Certainty of gray-level co-occurrence

ASM = \( {\sum}_{i,j}f{\left(i,j\right)}^2 \),

 Contrast

Intensity contrast between pixel and its neighbor

CON = \( {\sum}_{i,j}\left|i\right.-{\left.j\right|}^2f\left(i,j\right) \)

 Correlation

Linear gray-level dependence

COR = \( {\sum}_{i,j}\frac{{\left(i-{\mu}_i\right)}^{\ast }{\left(j-{\mu}_j\right)}^{\ast }f\left(i,j\right)}{\left(1+{\left(i-j\right)}^2\right)} \)

 Inverse difference moment (IDM)

Local homogeneity in gray-level

co-occurrence

IDM = \( {\sum}_{i,j}\frac{f\left(i,j\right)}{\left(1+{\left(i-j\right)}^2\right)} \)

 Entropy

Uncertainty of gray-level

co-occurrence

ENT = \( -{\sum}_{i,j}f{\left(i,j\right)}^{\ast}\log \left(f\left(i,j\right)\right) \)