These landscapes look really very well!
I especially like the 3rd one: It reminds me on Ayers Rock in Australia. 

I wonder how much of the detailed structure comes from the 
texture and how much from the topography (which, I guess, is 
an isosurface). 
[And of course, I am curious how it was done... are there chances that 
you publish some hints? :) ]
 
> Since this work seems rather boring to me i got the idea that one could
> have the computer automatically do this using standard optimization
> methods. 
>
I actually had a similar (but easier) experiment some time ago: 
It was about zooming into the Mandelbrot set (film-like) by having 
the computer find interesting regions by itself. 

In this case, the primary rating function was the variance of the 
chosen image regions. 

While this simple method is probably not applicable here, something 
like that could at least be used to sort out images we certainly do 
NOT want (by doing a 150x100 test trace). 

> So my idea is to render a set of images (maybe 10-20, something like a
> reasonable population size for an evolution strategy) in small size
> every week and have a public voting on them and use this vote for the
> ranking.  Then maybe render the highest rated image in larger, generate
> the next step population and start again.
> 
When visiting your homepage the last time, I wondered why there was the 
possibility to have the landscapes rated. Okay, now that becomes clear...

Wolfgang

Post a reply to this message