Full text: Proceedings of the international symposium on remote sensing for observation and inventory of earth resources and the endangered environment (Volume 1)

  
- 448 - 
board pattern, which has n-1 elements of the same size and 
one element, which has a size 4P-times larger than the size 
of one of the other elements. 
P is called perception value and is element of 2/4,3/4,4/4, 00 « 
n is the number of significant elementary object images in the 
image section Gp. 
4P must be always less n. 
The amount of area of the mask M must not be greater than 
Crea" ] £4. If F, calculated from the equation above,, 
is greater than Cre" y, then the according image 
Structure should not be called a texture. 
The maximum amount of the area of the mask is derived from the 
following demand: 
If a stripe structure is given, where the operator D is charac- 
terized by a sinus function, this stripe structure should be a 
texture, if there is a profile through Gg with at least one full 
sinus phase. 
Besides the two conditions T1 and T2 a third condition is necessa- 
ry for a image structure to be a texture: 
T3 If an image structure is texture for a perception value P1 
with a significance factor c, then the image structure will 
be a texture too for a perception value P1+1/4. 
If an image structure is texture for a significance factor c1 
with a percéption value P, then the image structure will be 
a texture too for each significance factor greater cl. 
If the conditions T1, T2 and T3 are satisfied, an image structure 
is called a texture with perception value P and significance fac- 
tor Ce 
The size of the mask M is calculated from shape parameters so that 
the definition of "texture" is derived from the shape parameters. 
So the shape parameters, whieh supply the "density variance" para- 
meters, to which the parameters of Haralick belong, are of great 
importance for characterizing textures. 
5. The significance of the texture definition given in 4. for 
image analysis in geosciences 
Before the shape parameters can be computed, the elementary ob- 
ject' images have to be filtered out by a method of matched fil- 
tering or by aid of equidensities. By smoothing the image before 
filtering the finest structures can be eliminated, if desired. 
Now analysing textures by aid of shape parameters is analysing 
properties and arrangement of significant elementary object ima- 
ges and the results of measurements of these properties can be 
transferred to the image significant objects on the earth surface. 
According as good the correlation between measurable properties 
of the significant elementary object images and the appropriate 
measurable properties of those objects on the earth surface, which 
are interesting for the interpreter, texture parameters can be 
measurable parameters for these objects. 
   
	        
Waiting...

Note to user

Dear user,

In response to current developments in the web technology used by the Goobi viewer, the software no longer supports your browser.

Please use one of the following browsers to display this page correctly.

Thank you.