Texture measures Group 2: Measures related to orderliness
Orderliness means: how regular (orderly) the pixel values are within the window.
Example: the two images below have the same degree of horizontal contrast (every pixel is one less than its eastern neighbour). But the degree of order is quite different:
| 1 | 2 | 3 | 4 | 3 | 4 | 3 | 2 | |
| 1 | 2 | 3 | 4 | 1 | 2 | 3 | 4 | |
| 1 | 2 | 3 | 4 | 2 | 3 | 4 | 5 | |
| 1 | 2 | 3 | 4 | 4 | 5 | 6 | 7 |
In the more orderly image on the left, each pair of values occurs many times: 2 is next to 1 four times, 3 is next to 2 four times, etc. For the less orderly image, combinations occur less often: 2 is next to 1 only once, 3 next to 2 three times, and so on.
Principle behind orderliness measures:
Orderliness measures, like contrast measures, use a weighted average of the GLCM values. The weight is constructed related to how many times a given pair occurs, so
- A weight that increases with commonness will yield a texture measure that increases with orderliness.
- A weight that decreases with commonness yields a texture measure that increases with disorder
Weights in orderliness measures:
Since the Pij values in the GLCM are already a measure of commonness of occurrence, it makes sense to use them in some form as weights for themselves.
Click on the links below for an explanation of each orderliness measure.
Click on next to go to ASM.