Welcome to Fractal Forums

Fractal Software => Programming => Topic started by: damith on June 08, 2010, 04:55:26 AM




Title: modeling trees using Iterated function system
Post by: damith on June 08, 2010, 04:55:26 AM
Hi,

I am currently involved in modeling of natural trees. I have used Honda's i-model for that purpose and ended up with good results. I would love to extend my work in this field. I have tried to model trees using L-systems but gave up due to complexity of algorithm. Has anyone interested or have done modelling trees? :embarrass:



Title: Re: modeling trees using Iterated function system
Post by: Lee Oliver on June 08, 2010, 06:19:59 AM
That's neat, the first fractal I posted was a tree ;D

I have found that Apophysis is a great program for making trees.  I've made multiple trees with this program and you really don't have to mess with any algorithms.

Here is my first tree :):

(http://www.fractalforums.com/gallery/1/1787_31_03_10_2_30_13.jpeg)


Title: Re: modeling trees using Iterated function system
Post by: damith on June 08, 2010, 09:58:52 AM
Hey BrutalToad. Nice work... :dink:

I have used MatLab to create the fractals. I have never used Apophysis. Thnx for the tip.  ^-^


Title: Re: modeling trees using Iterated function system
Post by: David Makin on June 08, 2010, 01:53:53 PM
Am working on a formula for Ultra Fractal to do this using escape-time LRRIFS (language restricted recurrent IFS), early results:

http://makinmagic.deviantart.com/art/Sierpinski-Tree-129956267?q=gallery%3AMakinMagic+sort%3Atime+in%3Ascraps&qo=11

http://makinmagic.deviantart.com/art/Spheres-130626934?q=gallery%3AMakinMagic+sort%3Atime+in%3Ascraps&qo=10

http://makinmagic.deviantart.com/art/Wooden-Frame-130687843?q=gallery%3AMakinMagic+sort%3Atime+in%3Ascraps&qo=9

I haven't released the code yet - I was working on it when Daniel White and Paul Nylander introduced us all to the Mandelbulb so working on the transformation of primitives got shelved - I will be going back to it eventually :)


Title: Re: modeling trees using Iterated function system
Post by: Bent-Winged Angel on June 08, 2010, 10:33:32 PM
I just got this done this morning.. BrutalTaod nice!

(http://www.fractalforums.com/gallery/2/1953_08_06_10_3_48_08.bmp)


Title: Re: modeling trees using Iterated function system
Post by: kram1032 on June 08, 2010, 10:57:27 PM
David:
I guess, with the newer discovered kaleidoscopic objects, this will get a boost in appearence?


Title: Re: modeling trees using Iterated function system
Post by: David Makin on June 09, 2010, 03:21:55 AM
David:
I guess, with the newer discovered kaleidoscopic objects, this will get a boost in appearence?

At the moment the code I was developing for IFS trees using primitives is using Hart's method for rendering because (for purely affine objects) that is magnitudes faster than using distance estimation, I suppose it's possible to do IFS trees using the distance estimation method but I wasn't thinking of trying that any time soon.
When I say "magnitudes faster" I'm talking about 10 to 20 times faster - for instance the latest (unreleased) version of my code using Hart's method renders the Menger Sponge @640*480 in 4.5 seconds flat on this ancient 3GHz P4HT (compared to over a minute for the Menger at the same size using DE methods).
I should add that even the released version of mmf4.ufm:3D IFS renders the default Menger in just 13.2 secs on this system.


Title: Re: modeling trees using Iterated function system
Post by: kram1032 on June 09, 2010, 03:00:59 PM
I see.
Well, I thought of generally possible shapes and not of speed :)


Title: Re: modeling trees using Iterated function system
Post by: David Makin on June 09, 2010, 11:10:47 PM
I see.
Well, I thought of generally possible shapes and not of speed :)

Ah, but I suspect your system is a little faster than mine ;)


Title: Re: modeling trees using Iterated function system
Post by: kram1032 on June 09, 2010, 11:22:20 PM
I'm not so sure about that...
I didn't find out what I did but something seriously slows down my computer. With the exact same specs, the exact same program usually performs better for others than it does for me...

I run a 3 GHz Pentium D and 1,25 GB RAM... There surely are better systems than that ;)


Title: Re: modeling trees using Iterated function system
Post by: David Makin on June 09, 2010, 11:49:02 PM
I'm not so sure about that...
I didn't find out what I did but something seriously slows down my computer. With the exact same specs, the exact same program usually performs better for others than it does for me...

I run a 3 GHz Pentium D and 1,25 GB RAM... There surely are better systems than that ;)

The 2GHz Pentium D (Dell Optiplex) we have at work is more than twice as fast as this 3GHz P4HT ;)