Hi all,
As my day job I am a lecturer in materials engineering at Swansea university in the UK. Within my research group two of my colleagues Prof. N. McMurray and Dr G. Williams have developed a machine that can be used for forensic detection of fingerprints on metal such as bullet casings or bomb fragments. The techniques is non contact so does not destroy evidence and can even show fingeprints that have apparently been "rubbed off" surfaces as it relies on the fact that when a person touches an object they leave behind minute traces of salt. This salt starts a corrosion reaction on the metal and our machine can detect these tiny corrosion voltages (billonths of a volt) and plot the data out to show fingerprints. This raw data is stored as a grid of these voltages so I thought that I could convert this into a heightfield and render the data in 3d. I initially did this using the landscape package vistapro and the images I produced were used in the UK version of FHM magazine and in the Daily Mail. I have now re-rendered some of this data using terragen 2 alpha and thought I would share them with you guys. Thanks to littlecannon for the basis of the canyon tgd.
Dr James Sullivan
This is awesome! Cool work!!! :o
Wow. That is pretty awesome. And the execution on TG2 is pretty neat as well :)
very good :)
except the grainy look of he clouds ^^
i am not a big fan of the clouds' shape too but the terrain is really good !
Fascinating!, as a watcher of 'The Crime Channel' :) and a terrainer, this is an unexpected pleasure.
monks
Interesting technology... sounds cool and useful.. :)
nice render!
Thanks for the comments. If you want some more info on how the technique works have a look at our website
http://www.swan.ac.uk/engineering/Research/MaterialsResearchCentre/ResearchAreas/CorrosionandCoatings/Forensicfingerprint/ (http://www.swan.ac.uk/engineering/Research/MaterialsResearchCentre/ResearchAreas/CorrosionandCoatings/Forensicfingerprint/)
We are hoping to develop this further with help from the UK forensic science service and the home office. It is good to see that presenting scientific data in this novel way has captured the interest of the public. A lot of our corrosion research produces data in this way so I will be using terragen 2 to render it where appropriate. It can be quite interesting to see that nature on a tiny scale can look very much like large scale terrains hence highlighting the fractal nature of the world around us.
James
This really is impressing.
It reminds me of something I tried when I was still using TG09:
Put something white or gray that has an interesting structure onto a scanner and you get a heightmap because the further a part is from the lightsource the darker it gets.
Works perfect with corals, bleached wood, golfballs, horseteeth...
If anyone is interested I can post my results, but it´s something anyone can try for himself.
Maybe I post it anyway, not anyone has a coral at hand.
Best regards,
Jan
@Mahnmut: would be interesting to see your results! ;)
AAARRRGGGHHH! :o
How did you get my fingerprints? Salt? Damn, I'm going down now...
Seriously though, nicely rendered, I especially love the banding and distribution of strata layer colours in the scene, very nice!
Quote from: dandelO on June 29, 2008, 05:41:06 PM
AAARRRGGGHHH! :o
How did you get my fingerprints? Salt? Damn, I'm going down now...
Seriously though, nicely rendered, I especially love the banding and distribution of strata layer colours in the scene, very nice!
:D
nice work its always nice to see new ideas ;)
The second image is cool. Quite dramatic and a feeling of big space. The spongy solid looking clouds spoils it a bit but it's not easy to get around that with this type of cumulous cloud.
Just a thought about why these cumulus clouds in TG2 never look right. If you look at real clouds of this type the light scatters through the clouds after entering them so in TG2 the tops of the clouds look OK but you still see forms underneath where the light should have been scattered and turned these parts to what looks like an even grey mass. You don't see those globular forms in real clouds. I've never tried strong cumulous shaped clouds so I don't know if anything can be done but we don't want to see that definition underneath. There should be no shadows contrasted with lit areas. The light is coming through scattered from inside the cloud above.
Efflux,
That definition you see in the shadows is caused by the Fake Internal Scattering. It can be disabled by setting that parameter to 0. In future versions I would like to make this fake internal scattering function apply differently to light coming from different directions which would make it more realistic.
Larger areas of scattering through the clouds should be handled by the Enviro Light and GI settings.
Matt
OK. Thanks Matt. I've not tried this type of cloud yet mainly due to impatience at rendering times but things are getting faster. Maybe others can experiment. I'm moving into having more time for this stuff now but working on the dark side with Mojoworld at the moment. Ideas in one app can feed back into the other though. I will eventually get around to trying clouds like this.