Dione bump map
Page 1 of 1

Author:  John Van Vliet [ Mon, 10-05-10, 5:26 GMT ]
Post subject:  Dione bump map

for any who have not been fallowing the UMSF thread ... topic=6543
-- and the one at shatters --
an update
this is only the higher res shots only 24 images
a 2k resize

all that has been done it to run Qtie on them with the PDS 23k map
then ran automos . This is a test only .

but it is getting there some are lining up well some need a net file ( well all need a control net )
--- test - the same image -- 600 px resize

no non clickable

Author:  John Van Vliet [ Sun, 30-05-10, 4:55 GMT ]
Post subject:  re

just an update
a 4k resize of the " work in progress " image

still needs a bunch of work

Author:  t00fri [ Sun, 30-05-10, 10:23 GMT ]
Post subject: 

Hi John,

thanks for linking this very interesting UMSF thread about bumpmap algorithms derived from shading into CM!

For me it still seems like a bit of a "miracle", since at least partly, genuine elevation/bump measurements imply spacial information that is INDEPENDENT from shaded basemaps. I can believe that shaded basemaps allow for some amount of bumpmap reconstruction via certain clever algorithms, since the shading does display some 3d aspects of a surface. I can also believe that shaded basemaps with light coming from a multitude of directions might allow for virtually complete bumpmap reconstructions. However, we typically have shaded maps (i.e. photos) available only for light from ONE direction, so a large amount of the bumpmap must remain effectively unknown with this reconstruction technique ...

Did I forget something essential in my argument?

I believe a more mathematically rigorous argumentation can be done by means of Fourier transformations.


Author:  John Van Vliet [ Sun, 30-05-10, 18:02 GMT ]
Post subject:  re

at the moment the map has only limited scientific info.( almost none )
to salve two issues i need to toss out a lot of info

but it can make a map that "looks nice"
and has some relative height info

someone is working on something better , but it will be many,many years before it is able to be used by the general public ... topic=6413
I believe a more mathematically rigorous argumentation can be done by means of Fourier transformations.

a Fourier transformations is on the right and an image from cyclops is on the left


Author:  John Van Vliet [ Mon, 07-06-10, 21:16 GMT ]
Post subject:  re

have you read Bjorn Jonsson recent post on the thread
-- excerpt --
After fixing it I'm getting very promising results but it is obvious that the best results are going to come from combining several approaches: Stereo, Cyclops and my primitive SFS software.

in this i am in agreement stereo pairs and sfs is needed along with everything else .

and a artistic eye for a clean and workable map

for right know this dose work nicely for a small area , and a lot harder for a whole map

Page 1 of 1 All times are UTC
Powered by phpBB® Forum Software © phpBB Group