POV-Ray : Newsgroups : povray.binaries.images : Fibonacci Mosaics (151KB & 145KB) Server Time
28 Mar 2024 17:06:26 EDT (-0400)
  Fibonacci Mosaics (151KB & 145KB) (Message 15 to 24 of 24)  
<<< Previous 10 Messages Goto Initial 10 Messages
From: Ron Parker
Subject: Re: Fibonacci Mosaics (151KB & 145KB) - crackle.jpg (1/1)
Date: 8 Sep 2000 10:44:35
Message: <slrn8rhvj6.19a.ron.parker@fwi.com>
On Thu, 07 Sep 2000 19:39:16 -0500, Chris Huff wrote:
>I have attached an image of a sphere with a crackle pigment. As you see, 
>it has separate "cells". In the official version and MegaPOV, these are 
>located randomly, I was wondering if it would be possible to specify 
>their locations and strengths manually, maybe as an extension to the 
>blob pattern.

Yes, but you'd need an efficient data structure to find the closest center
to a given sample point quickly.  Fortunately, I've done some work in that 
field (though in 2D) so I have some ideas on how it might be done.

Such a pattern would solve Gilles' lizard-skin problem, too.

-- 
Ron Parker   http://www2.fwi.com/~parkerr/traces.html
My opinions.  Mine.  Not anyone else's.


Post a reply to this message

From: Ron Parker
Subject: Re: Fibonacci Mosaics (151KB & 145KB)
Date: 8 Sep 2000 10:47:40
Message: <slrn8rhvp0.19a.ron.parker@fwi.com>
On Thu, 07 Sep 2000 22:12:32 -0500, David Fontaine wrote:
>Tor Olav Kristensen wrote:
>
>> It's also interesting to see how well the curves
>> in the Fibonacci pattern follows the edges of the
>> flower's leafs.
>
>Yeah, there's probably 1000 different places you could find the golden
>ratio or a Fibannochi number on a single plant...

You realize, of course, that there's a relationship between the two.

If F(n) is the n'th Fibonacci number, then 

  lim    F(n)/F(n-1) = phi      
n -> oo 

(where "oo" should be taken as a low-resolution version of the symbol for
infinity.)

-- 
Ron Parker   http://www2.fwi.com/~parkerr/traces.html
My opinions.  Mine.  Not anyone else's.


Post a reply to this message

From: Alf Peake
Subject: Re: Fibonacci Mosaics (151KB & 145KB)
Date: 8 Sep 2000 18:30:30
Message: <39b96886@news.povray.org>
Tor Olav Kristensen <tto### [at] onlineno> wrote in message
news:39B65D7C.9BE96D79@online.no...
>
> Any suggestions for improvements ?
>
You've got to be joking?
:)

Alf


Post a reply to this message

From: Alf Peake
Subject: Re: Fibonacci Mosaics (151KB & 145KB)
Date: 8 Sep 2000 18:30:55
Message: <39b9689f@news.povray.org>
I think the pattern adds to these images. A regular 50*50 grid might
look good, but these 2 images say "Hey! Look at this!"

Alf


David Fontaine <dav### [at] faricynet> wrote in message
news:39B6C977.BCA27C4F@faricy.net...
> These are good! The first is a better subject for the process IMHO,
int the
> second the pattern is very distracting...
>
> --
> David Fontaine   <dav### [at] faricynet>   ICQ 55354965
> Please visit my website:  http://davidf.faricy.net/
>
>


Post a reply to this message

From: David Fontaine
Subject: Re: Fibonacci Mosaics (151KB & 145KB)
Date: 8 Sep 2000 20:39:30
Message: <39B985A9.968B832F@faricy.net>
Ron Parker wrote:

> You realize, of course, that there's a relationship between the two.

Well Duh

> If F(n) is the n'th Fibonacci number, then
>
>   lim    F(n)/F(n-1) = phi
> n -> oo

Well Duh

> (where "oo" should be taken as a low-resolution version of the symbol for
> infinity.)

Well Duh

:-)


teehee

--
David Fontaine   <dav### [at] faricynet>   ICQ 55354965
Please visit my website:  http://davidf.faricy.net/


Post a reply to this message

From: Chris S 
Subject: Re: Fibonacci Mosaics (151KB & 145KB)
Date: 8 Sep 2000 22:17:15
Message: <39b99dab$1@news.povray.org>
Tor Olav Kristensen <tto### [at] onlineno> wrote in message
news:39B8344B.27A063A3@online.no...
> I don't know how Frank Frazetta look(ed ?), but
> if it is showing him, then it must be a self-portrait.
>
> Does anyone know more about this ?

Yeah, if you take a look at the official Frazetta website
http://www.frazettaartgallery.com/ they mention this self-portrait briefly
in his bio.

-Chris-


Post a reply to this message

From: Ian Witham
Subject: Re: Fibonacci Mosaics (151KB & 145KB)
Date: 9 Sep 2000 09:43:05
Message: <39ba3e69@news.povray.org>
> Any suggestions for improvements ?

I've made some code which only parses the fibonacci antiblobs needed to make
any given rectangle rather than a circle....  It would have save a *lot* of
parsing time for these images and it would have looked better.  I can post
it if you"re interested ;-)


Post a reply to this message

From: Tor Olav Kristensen
Subject: Re: Fibonacci Mosaics (151KB & 145KB)
Date: 11 Sep 2000 21:05:12
Message: <39BD8000.BA4271C3@online.no>
Ian Witham wrote:

> > Any suggestions for improvements ?
>
> I've made some code which only parses the fibonacci antiblobs needed to make
> any given rectangle rather than a circle....  It would have save a *lot* of
> parsing time for these images and it would have looked better.  I can post
> it if you"re interested ;-)

Yes, please.


Tor Olav
--
mailto:tor### [at] hotmailcom
http://www.crosswinds.net/~tok/tokrays.html


Post a reply to this message

From: Chris Huff
Subject: Re: Fibonacci Mosaics (151KB & 145KB) - crackle.jpg (1/1)
Date: 14 Sep 2000 08:07:55
Message: <chrishuff-F320F1.07095014092000@news.povray.org>


Post a reply to this message

From: Ron Parker
Subject: Re: Fibonacci Mosaics (151KB & 145KB) - crackle.jpg (1/1)
Date: 14 Sep 2000 10:51:12
Message: <slrn8s1q53.244.ron.parker@fwi.com>
On Thu, 14 Sep 2000 07:09:50 -0500, Chris Huff wrote:
>In article <slr### [at] fwicom>, ron### [at] povrayorg 
>wrote:
>
>> Yes, but you'd need an efficient data structure to find the closest 
>> center to a given sample point quickly.  Fortunately, I've done some 
>> work in that field (though in 2D) so I have some ideas on how it 
>> might be done.
>
>Any suggestions? I am thinking of it as an addition to the blob pattern, 
>which has several more component types than the blob object. Could this 
>data structure also handle line(cylinder) components? Of course, just 
>having it available for the spherical components(and components that can 
>be approximated by spheres) would help...

The code I had (in 2d) used a quadtree to store the points, and another to
store the lines.  Points are simple: find the quadtree node that contains
the sample point, and search it for closer points.  If any edges are closer
than the closest point, search the quadtree node on that side for closer
points as well.  If any corners are closer, search the node diagonally
adjacent to the current node as well.  Continue for all edges and corners
of all tested nodes, until none are closer than the closest point found.
(This is easier than it looks, due to the structure of a quadtree.)  For 3d,
of course, you'd use an octree, and you'd generalize the above description 
to faces, edges, and corners of an octree node.

To store lines, you have to corrupt the quadtree structure a bit by storing
lines that are too long to fit entirely in a leaf node in the interior nodes
of the tree.  The procedure is the same, though, with the caveat that you must
also search all ancestor nodes of any leaf node you search.  This structure 
is not as efficient as the one for points unless your line segments are all 
very small, with the possible exception of a few large ones.  In my case,
working with digital cartography, this was true.  In the case of a blob, it
may not be.  You may need to modify the criteria you use to split (for example,
it may make more sense to break the nodes on one axis at a time rather than 
all at once.)

If you generalize to 3d, it should become obvious that the same strategy
that works for line segments will work for polygons.

-- 
Ron Parker   http://www2.fwi.com/~parkerr/traces.html
My opinions.  Mine.  Not anyone else's.


Post a reply to this message

<<< Previous 10 Messages Goto Initial 10 Messages

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.