Haralick Texture
| Line 5: | Line 5: | ||
of the digital numbers within a window. An essential component of the concept of | of the digital numbers within a window. An essential component of the concept of | ||
the Haralick texture measures is the definition of eight nearest-neighbor resolution cells | the Haralick texture measures is the definition of eight nearest-neighbor resolution cells | ||
| − | (Fig.). Now we may define different matrices for different angles ( | + | (Fig.). Now we may define different matrices for different angles (0°,45°,90°,135°) and |
distances between the horizontal neighboring pixels. | distances between the horizontal neighboring pixels. | ||
Many studies in land cover and forest type classification utilize textural features to | Many studies in land cover and forest type classification utilize textural features to | ||
| − | improve the | + | improve the classification accuracies. |
We use the ASTER satellite band number 1 (green) and an inter-pixel sampling distance | We use the ASTER satellite band number 1 (green) and an inter-pixel sampling distance | ||
| − | of one. At first we need to linearly transform the 16-bit ASTER data to 8-bit | + | of one. At first we need to linearly transform the 16-bit ASTER data to 8-bit radiometric |
resolution. | resolution. | ||
| − | [[File:Texture.png|center|500px|thumb|3x3 window definition and spatial relationship for calculating Haralick texture measures. Pixel 1 and 5 are | + | [[File:Texture.png|center|500px|thumb|3x3 window definition and spatial relationship for calculating Haralick texture measures. Pixel 1 and 5 are 0° (horizontal) nearest neighbors to the center pixel * ; pixel 2 and 6 are 135° nearest neighbors; pixels 3 and 7 are 90° nearest neighbors, pixel 4 and 8 are 45° nearest neighbors to the center pixel * (Haralick et al. 1973)]] |
{{Exercise|message=Exercise 35|text=}} | {{Exercise|message=Exercise 35|text=}} | ||
Revision as of 16:27, 2 January 2011
Image texture is a quantification of the spatial variation of grey tone values. Haralick et al. (1973) presented texture measures that may be derived by comparing the values of the digital numbers within a window. An essential component of the concept of the Haralick texture measures is the definition of eight nearest-neighbor resolution cells (Fig.). Now we may define different matrices for different angles (0°,45°,90°,135°) and distances between the horizontal neighboring pixels. Many studies in land cover and forest type classification utilize textural features to improve the classification accuracies. We use the ASTER satellite band number 1 (green) and an inter-pixel sampling distance of one. At first we need to linearly transform the 16-bit ASTER data to 8-bit radiometric resolution.
