IPB

Welcome Guest ( Log In | Register )

High-Res DEMs from single HiRISE images, First results of new "Shape from Shading" algorithm
Nirgal
post Jan 16 2010, 03:30 PM
Post #1


Member
***

Group: Members
Posts: 713
Joined: 30-March 05
Member No.: 223



Hi all,

Here the long overdue continuation of the "Alien Landscapes" series. This time based on 3D
DEMs generated with "Shape from Shading" from single HiRISE images.
Enjoy smile.gif

Click on Images for larger version.

Detail views from PSP_002172_1410 (large gully system)





Detail view of Gullies from PSP_001376_1675





Detail of gully system in PSP_002022_1455



Dune Views from PSP_004339_1890





Detail from PSP_001834_1605





Here is some background info on the making of the images:

"Shape from Shading" (SFS) i.e. the possibility to extract shape information from a single image has always been a fascinating topic for me.
Now I found the time to implement a prototype for a new SFS algorithm based on some ideas that I've been thinking about for a long time.
The problem with existing SFS approaches (see here for a survey is that they either tend to over-smooth the details (due to the regularization constraint) or suffer from excessive noise in the high-frequency components of the reconstructed surface. Another problem is the large demand on CPU ressources which would make them very challenging to apply to large scale input data, such as HiRISE orbiter images.

So for a long time I was rather sceptical as to the potential of SFS and it was my impression that Methods based on multiple images (stereo) must be far superior to single-image SFS.

However, after a long time of experimenting, combining existing approaches with some new ideas, I got the following quite promising first results that I'd like to share:

All of the images were generated from a single HiRISE image (no depth information was used from stereo or laser altimeter data).
Also, no texturing or additional coloring/shading was applied when rendering the surface.
Every detail visible is real 3D down to the pixel-level...
For rendering I used a very simple model based on lambertian reflection with gouraud shading.

The resolution of the images is still moderate: that is downsampled details crops in the order of 0.5-1 Megapixels.
However, despite the heavy math machinery that drives the core of the algoritm (several systems of equations with millions of unknowns)
the processing time is still moderate (about 15 Minutes per med-res image, using about 2 Gigs main mem) such that the application to full-res HiRISE images should be possible smile.gif

The following image shows an example to illustrate the general principle (click to enlarge).


On the left hand side the 2D input image (simple noisy JPEG from the Web with unknwon light source direction). On the right hand side shows the recovered 3D surface re-lighted under a different light source direction.
Note that one problem of the current implementation of the algorithm is it's vulnerability to notable distortions in the low frequency components (i.e. large scale variations) of the generated surface. However I'm confident that this can be overcome by an improved version or by adding the large-scale depth information from stereo-based DEMs or altimeter data (MOLA) where available.
Go to the top of the page
 
+Quote Post
 
Start new topic
Replies
djellison
post Jan 19 2010, 04:19 PM
Post #2


Administrator
****

Group: Chairman
Posts: 13754
Joined: 8-February 04
Member No.: 1



OK - yeah - do you think you'll be able to share the software and technique at some point? I have enough ideas for this that would keep you on your toes till April smile.gif
Go to the top of the page
 
+Quote Post
Bjorn Jonsson
post Jan 19 2010, 05:37 PM
Post #3


IMG to PNG GOD
****

Group: Moderator
Posts: 1492
Joined: 19-February 04
From: Near fire and ice
Member No.: 38



QUOTE (Nirgal @ Jan 19 2010, 03:43 PM) *
ok, no problem: here I generated some views of it in 3D smile.gif


*Jaw drops to the floor* wink.gif

This is awesome, probably the most interesting UMSF-member-developed software I have seen in the history of UMSF so I can simply repeat what Doug said above (actually I could probably keep you busy till April *next* year or longer ;-). If I could combine output from SFS with the output from my steromatcher I'd get DEMs of probably something like 10 times higher resolution than I have ever dreamt of. Combining DEMs like this is probably not difficult. If you manage to get this to work the way you want to there are lots of people here who would love to test it.

And I would make my stereomatcher available - would need to spruce it up a lot though and it still has some bugs and quirks. Combining results from these two approaches opens up *lots* of possibilities. In particular, the low frequency variations from my software are pretty accurate (how accurate depends largely on how accurate the viewing geometry information is) while high frequency details are a problem and frequently 'disappear'.

QUOTE (Nirgal @ Jan 19 2010, 03:43 PM) *
P.S.: one nice thing about the algorithm is that one does not need to specify any external calibration parameters, like the light source (sun incidence and azimut) direction.( because those are estimated simultaneously with the recoverered 3D surface )

In contrast my stereomatcher needs accurate viewing geometry information, field of view etc. Inaccurate viewing geometry information typically manifests itself as a 'tilted' DEM.
Go to the top of the page
 
+Quote Post

Posts in this topic
- Nirgal   High-Res DEMs from single HiRISE images   Jan 16 2010, 03:30 PM
- - lyford   WOAH! I want to climb those dunes..... am...   Jan 16 2010, 04:29 PM
- - mhoward   Mind-blowing.   Jan 16 2010, 05:03 PM
- - cbcnasa   The images are just amazing, fantastic job.   Jan 16 2010, 05:48 PM
- - jamescanvin   Those are just astonishing Bernhard, great work...   Jan 16 2010, 06:05 PM
- - ElkGroveDan   Amazing and beautiful stuff, as always. I don...   Jan 16 2010, 06:39 PM
- - eoincampbell   Another UMSF classic, it's truly a joy to come...   Jan 16 2010, 07:13 PM
- - PDP8E   Bernhard, This is fabulous progress report! I...   Jan 16 2010, 07:43 PM
- - djellison   O M F G   Jan 16 2010, 08:20 PM
- - Tesheiner   Unbelievable!!! I want to climb one...   Jan 16 2010, 08:36 PM
- - Bjorn Jonsson   In a word: WOW!!! Are you draping th...   Jan 16 2010, 09:09 PM
|- - Nirgal   QUOTE (Bjorn Jonsson @ Jan 16 2010, 10:09...   Jan 16 2010, 10:27 PM
|- - Stu   QUOTE (Nirgal @ Jan 16 2010, 10:27 PM) P....   Jan 16 2010, 10:29 PM
|- - Bjorn Jonsson   QUOTE (Nirgal @ Jan 16 2010, 10:27 PM) Th...   Jan 16 2010, 11:07 PM
|- - Nirgal   QUOTE (Bjorn Jonsson @ Jan 17 2010, 12:07...   Jan 16 2010, 11:16 PM
|- - ngunn   QUOTE (Nirgal @ Jan 16 2010, 11:16 PM) no...   Jan 17 2010, 10:18 AM
||- - Nirgal   QUOTE (ngunn @ Jan 17 2010, 11:18 AM) Exc...   Jan 17 2010, 12:25 PM
|- - ElkGroveDan   And for an Earthly analog here is the Emperor Face...   Jan 17 2010, 06:09 PM
|- - Nirgal   QUOTE (ElkGroveDan @ Jan 17 2010, 07:09 P...   Jan 18 2010, 01:15 PM
- - Reckless   Brilliant Bernhard As I was looking at those pic...   Jan 16 2010, 09:25 PM
- - nprev   Just freakin' amazing, is all! What immedi...   Jan 16 2010, 09:32 PM
- - ngunn   Me too. I always prefer to see landscapes without ...   Jan 16 2010, 10:49 PM
- - nprev   ...how sedimentary can you get?   Jan 16 2010, 11:26 PM
- - djellison   If you want a good test - try the Pathfinder obser...   Jan 17 2010, 03:14 PM
- - imipak   ...what everyone else has said! I know very l...   Jan 17 2010, 07:17 PM
|- - Nirgal   QUOTE (imipak @ Jan 17 2010, 08:17 PM) .....   Jan 18 2010, 10:26 AM
|- - sgendreau   QUOTE (Nirgal @ Jan 18 2010, 02:26 AM) I ...   Jan 18 2010, 04:27 PM
- - vikingmars   Bernhard, your work is "estupendo" ...   Jan 18 2010, 02:49 PM
- - climber   Bernhard, after Olivier's statement, knowing w...   Jan 18 2010, 05:20 PM
- - PDP8E   Bernhard, What could your technique do with a fair...   Jan 18 2010, 08:42 PM
- - Bjorn Jonsson   Earlier in this thread I mentioned Enceladus and E...   Jan 19 2010, 01:00 AM
|- - Nirgal   QUOTE (Bjorn Jonsson @ Jan 19 2010, 02:00...   Jan 19 2010, 03:43 PM
- - djellison   OK - yeah - do you think you'll be able to sha...   Jan 19 2010, 04:19 PM
|- - Bjorn Jonsson   QUOTE (Nirgal @ Jan 19 2010, 03:43 PM) ok...   Jan 19 2010, 05:37 PM
- - djellison   Things I want to try ( in no particular order ) T...   Jan 19 2010, 05:47 PM
|- - vikingmars   QUOTE (djellison @ Jan 19 2010, 06:47 PM)...   Jan 20 2010, 08:37 AM
- - mcaplinger   Not to be dismissive of this, but as Nirgal pointe...   Jan 19 2010, 06:23 PM
|- - Nirgal   QUOTE (mcaplinger @ Jan 19 2010, 07:23 PM...   Jan 19 2010, 09:27 PM
- - ElkGroveDan   Since we have a lot of examples of places with ver...   Jan 19 2010, 06:38 PM
- - Ant103   ASTOUNDING ! AMAZING ! You just did a WOND...   Jan 19 2010, 09:30 PM
- - Stu   You keep sharing your "pretty visualisations...   Jan 19 2010, 09:35 PM
- - Bjorn Jonsson   There are lots and lots of things I'd be inter...   Jan 19 2010, 11:05 PM
- - 4th rock from the sun   I'd VERY much like to produce a detailed DEM f...   Jan 20 2010, 10:09 AM
|- - mcaplinger   QUOTE (4th rock from the sun @ Jan 20 2010, 0...   Jan 20 2010, 02:19 PM
- - djellison   I'm not sure this would work on Radar.   Jan 20 2010, 01:56 PM
|- - 4th rock from the sun   It should work on radar, as there's the concep...   Jan 20 2010, 02:13 PM
- - imipak   *cough* Titan..? (Edit - sorry, I didn't spot ...   Jan 20 2010, 02:42 PM
- - elakdawalla   The software wouldn't work on radar images wit...   Jan 20 2010, 07:30 PM
|- - mcaplinger   QUOTE (elakdawalla @ Jan 20 2010, 11:30 A...   Jan 20 2010, 07:56 PM
- - vikingmars   Bernhard, just an idea... Because this 3D HI-Res D...   Jan 20 2010, 08:38 PM
|- - Nirgal   QUOTE (vikingmars @ Jan 20 2010, 09:38 PM...   Jan 20 2010, 09:30 PM
|- - Bjorn Jonsson   As should already be obvious I'm very interest...   Jan 21 2010, 12:56 AM
- - 4th rock from the sun   Thanks for sharing your insights into the photocli...   Jan 20 2010, 08:43 PM
- - climber   I'm wondering if you (Bernhard) can run your s...   Jan 21 2010, 10:07 PM
- - Phil Stooke   It can't be done on that image, it would have ...   Jan 22 2010, 12:32 PM
|- - Nirgal   QUOTE (Phil Stooke @ Jan 22 2010, 01:32 P...   Jan 22 2010, 03:40 PM
- - djellison   Two for the columbia hills are listed within this ...   Jan 22 2010, 03:59 PM
|- - Nirgal   QUOTE (djellison @ Jan 22 2010, 04:59 PM)...   Jan 23 2010, 04:00 PM
- - ngunn   Nirgal, I can't find your original thread so I...   Mar 1 2010, 10:08 PM
|- - Nirgal   QUOTE (ngunn @ Mar 1 2010, 11:08 PM) Nirg...   Mar 1 2010, 10:36 PM
- - ngunn   No I don't, unfortunately. Go to the 'Ende...   Mar 1 2010, 10:57 PM
- - Tesheiner   AFAIK, that's the non-map-projected version of...   Mar 2 2010, 06:52 AM
- - Nirgal   Upon a request posed over at the "Dust Avalan...   Mar 2 2010, 08:44 PM
|- - tdemko   This area has fascinated me for a while and Nirgal...   Mar 3 2010, 11:15 AM
|- - ngunn   QUOTE (tdemko @ Mar 3 2010, 11:15 AM) The...   Mar 3 2010, 12:57 PM
||- - Nirgal   QUOTE (ngunn @ Mar 3 2010, 01:57 PM) Fasc...   Mar 3 2010, 03:27 PM
|- - ElkGroveDan   QUOTE (tdemko @ Mar 3 2010, 03:15 AM) Thi...   Mar 3 2010, 02:32 PM
- - ngunn   YES!! Now that is everything I'd hoped...   Mar 2 2010, 09:03 PM
|- - Nirgal   QUOTE (ngunn @ Mar 2 2010, 10:03 PM) Can ...   Mar 2 2010, 09:33 PM
- - Tesheiner   Incredible! BTW, I think we'll have to sa...   Mar 2 2010, 09:06 PM
- - Floyd   A question for Tim and Nirgal: This is one of the...   Mar 3 2010, 02:33 PM
- - ngunn   Zoomable CTX of the area: http://global-data.mars....   Mar 3 2010, 02:58 PM
|- - Nirgal   QUOTE (ngunn @ Mar 3 2010, 03:58 PM) Look...   Mar 4 2010, 08:06 AM
- - Nirgal   For comparison, here is (a crop of) the original, ...   Mar 3 2010, 06:37 PM
- - Julius   To me it looks like an' island' which even...   Mar 3 2010, 07:11 PM
- - ngunn   Again, greatly appreciated Nirgal. Those ones show...   Mar 4 2010, 09:26 AM
- - ngunn   Nirgal, on closer inspection I'm having troubl...   Mar 4 2010, 02:19 PM
- - Nirgal   QUOTE (ngunn @ Mar 4 2010, 03:19 PM) Nirg...   Mar 4 2010, 03:57 PM


Reply to this topicStart new topic

 



RSS Lo-Fi Version Time is now: 16th September 2014 - 07:30 PM
RULES AND GUIDELINES
Please read the Forum Rules and Guidelines before posting.

IMAGE COPYRIGHT
Images posted on UnmannedSpaceflight.com may be copyrighted. Do not reproduce without permission. Read here for further information on space images and copyright.

OPINIONS AND MODERATION
Opinions expressed on UnmannedSpaceflight.com are those of the individual posters and do not necessarily reflect the opinions of UnmannedSpaceflight.com or The Planetary Society. The all-volunteer UnmannedSpaceflight.com moderation team is wholly independent of The Planetary Society. The Planetary Society has no influence over decisions made by the UnmannedSpaceflight.com moderators.
SUPPORT THE FORUM
Unmannedspaceflight.com is a project of the Planetary Society and is funded by donations from visitors and members. Help keep this forum up and running by contributing here.