END OF AN ERA, FRACTALFORUMS.COM IS CONTINUED ON FRACTALFORUMS.ORG

I was messing with fourier transform and wavelet decompose.scm plugins for the Gimp image editing software to detect zones of different frequency levels in color fractal pics online in an attempt to extract the dimension gradient information from fractal pics and rebuild it into heightmaps where the small details are white, like the tops of mountains, and the lesser details are grey shades of the heightmaps and the low details form dark blobs underneath the lighter colors that say where the main masses of the mountains are.

I think i am confusing myself wether it's possible to get frequency data from an image and rebuild it into coherent heightmaps for fractals. is it possible? It's deceptively abstract, and i am sure some of you have attempted similar maths, please share your experiences. The infinity border of the fractal has very high dynamic range between dark and light, which means that the middle of the heightmap will normally have noise pixels where the mountain goes from high to low in the space of one pixel, and with wavelet decomposition we should be able to isolate the high frequencies, set their areas dark and substract all the angle traps from the rest so that they effectively add only 2% altitude to the rest of the fractal curves, as they only vary over 2% of the pic. haven't totally managed that still. It woud be cool to have a filter on the thing called : heightmap a fractal pic! and it would take out the dynamic range on the small details.

I've been messing with HGT2STL program by DWilbourne and figuring out various ways of making heightmaps from zoom pics from the net. here are some images: Be warned, HGT2STL makes 2gb mesh files from 2048x2048 pics so you need 64 bit version of meshlab and alot of ram in order to make pro nice 3d mesh. Please if you have a 64 bit PC and lots of RAM i can send you some great fractals if you can return me some refined mesh from meshlab