topographic

topographic_sketch ( Image : Sketch : : )

Compute the topographic primal sketch of an image.

topographic_sketch computes the topographic primal sketch of the input image Image. This is done by approximating the image locally by a bicubic polynomial (``facet model''). It serves to calculate the first and second partial derivatives of the image, and thus to classify the image into 11 classes. These classes are coded in the output image Sketch as numbers from 1 to 11. The classes are as follows:

Peak                  1
Pit                   2
Ridge                 3
Ravine                4
Saddle                5
Flat                  6
Hillside Slope        7
Hillside Convex       8
Hillside Concave      9
Hillside Saddle      10
Hillside Inflection  11

In order to obtain the separate classes as regions, a threshold operation has to be applied to the result image with the appropriate thresholds.

Parameters

`Image` (input_object)	(multichannel-)image(-array) `->` object : byte
Image for which the topographic primal sketch is to be computed.

`Sketch` (output_object)	(multichannel-)image(-array) `->` object : byte
Label image containing the 11 classes.

Example

/* To extract the Ridges from a Image */
read_image(Image,'sinus') 
topographic_sketch(Image,Sketch) 
threshold(Sketch,Ridges,3,3).

Complexity

Let n be the number of pixels in the image. Then O(n) operations are performed.

Result

topographic_sketch returns 2 (H_MSG_TRUE) if all parameters are correct. If the input is empty the behaviour can be set via set_system('no_object_result',<Result>). If necessary, an exception is raised.

Parallelization Information

topographic_sketch is reentrant and automatically parallelized (on tuple level, channel level).

Possible Successors

threshold

References

R. Haralick, L. Shapiro: ``Computer and Robot Vision, Volume I''; Reading, Massachusetts, Addison-Wesley; 1992; Kapitel 8.13.

Module

Image filters