Welcome to Fractal Forums

Fractal Math, Chaos Theory & Research => (new) Theories & Research => Topic started by: cKleinhuis on April 29, 2010, 07:29:43 PM




Title: My Bachelor Thesis ? Fractal Mountains revisited
Post by: cKleinhuis on April 29, 2010, 07:29:43 PM
hi there, i planning to write my bachelor thesis about this method, invented by me, the thesis will deal with pro and cons of the method,
here a little peak, and i am very intrested on your opinion on this thesis:

we all know the classig BERG algorithm to create mountain scapes, here demonstrated in 2d:
1. take the midpoint of a line and move it randomly a certain amount up or down
2. repeat with each new created line with step 1

the result is a mountain or hilly like shape:


http://digitalgott.de/transfer/bachelor/mountains_8iterations.swf
(click on image to regenerate)


but the above image has some serious drawbacks, so it has no right angles, caves or advances
so, i have a slight modification to the original algorithm:

1. take the midpoint of a line and move it by a random amount in the direction of the normal
2. repeat with each new created line with step 1

here is a 2d implementation of this method, but my work aims to use it on 3d planes ...

http://digitalgott.de/transfer/bachelor/mountains_10iterations.swf
(click on image to regenerate)



what do you think ?

i need to convience my prof ;)



Title: Re: My Bachelor Thesis: Fractal Mountains revisited
Post by: bib on April 29, 2010, 07:34:04 PM
My first impression is that the second one looks less realistic than the first one, because of the horizontal small peaks. Maybe a combination of both would do better?...

Or instead of using the normal between 2 points, use the normal between 2 more distant points, so that makes a kind of average.


Title: Re: My Bachelor Thesis: Fractal Mountains revisited
Post by: reesej2 on April 29, 2010, 07:37:59 PM
The second image looks a little "squashed" but I'd guess that's just a matter of calibrating the parameters. You may want to make the horizontal distortions less common. I'd like to see a 3D rendering.... maybe I'll give it a shot myself.

What were you planning on discussing in your thesis? A mathematical perspective (fractal dimension, correlation to real mountains, etc) or a computational one (order of complexity, ease of rendering, etc)? Or some combination of the two?


Title: Re: My Bachelor Thesis: Fractal Mountains revisited
Post by: cKleinhuis on April 29, 2010, 07:50:38 PM
as a bachelor thesis i think it will contain an implementation and a comparison to classic method, but i dont know yet, it is just one of the themes i would actually love to write about ;)

the parameters indeed must be tweaked a little, but i believe the second one has after a few sure surrealistic formations a better outline of a mountainvies, e.g. compare it to a foto:
(http://www.weltum.de/weltum/img/alpen_matterhorn.jpg)
in that image are also small advances, and those can not be displayed using a height map

but i agree that there are to  sharp gaps, this should be handled by an angles/max value

my personal aim is to generate visually more attractive computer generated landscapes ;)


Title: Re: My Bachelor Thesis ? Fractal Mountains revisited
Post by: Nahee_Enterprises on April 30, 2010, 01:23:25 PM
The first example seems more realistic to me than the second one.

And I found it interesting that I could click anywhere inside of the second one to get it to create another pattern.  But I could only click within the gray "mountain" area of the first one, because the white "sky" area would not cause a new pattern to be generated.