3D Modeling Existing Environments

Become a member of the CGSociety

Connect, Share, and Learn with our Large Growing CG Art Community. It's Free!

THREAD CLOSED
 
Thread Tools Search this Thread Display Modes
Old 01 January 2013   #16
Originally Posted by AJ1: Yea, Google doesn't like people doing it, and it looks like its been bumped from the search results. The software was called OGLE, and the site seems to be down.


-AJ


yo can get access to the ogle site via wayback machine

http://web.archive.org/web/20100709...amresearch.org/
__________________
popol va au cirque ....
 
Old 01 January 2013   #17
Originally Posted by VisitorfromArea51: Hey Manuel,

Concerning the camera calibration and point cloud efforts, can you recall the amount of manpower / time it took (days, weeks, months)?


Hi Brian,

well if I was to say a number it'll be an arbitrary number, because imo that depends on 3 things mainly . Quality of the images provided, size of your team and number of hours on the task, not mentioning expertise on the matter...

So disregarding that. I'll say that it is a really straight forward technique, sure as anything you can get fancy and enhance a lot of stuff in compositing, including exporting 3d models done in photogrammetry to nuke for further enhancement with textures and all... also doing all the shading work per material per object... create different maps for each...hdri creation based on the environment... etc etc...!!!

But photogrammetry by itself is quite simple, you need a minimum of 3 images per object to be modeled (..at least in the soft I used), those images does not need to have the exact same lens settings, but you need to know which image and which lens setting was used on each photo, so that, at the time of calibration the photo sources will be interpreted correctly.

Once the point cloud is generated (in your calibration soft of choice) you will need to export that information (taking always in consideration the metric system and scale you used, and make sure that your 3d app is set up accordingly) you can also create some dummy place holder geo inside the calibration package for visual help.

Then with the point cloud inside the 3d package, you will setup each camera as on your 3d calibration software, and assign an image to each. Once this is done, you will be able to model accurately! Then you will use the same photos at a higher resolution to create textures for the buildings/props....etc with multiple camera projections... There are many more details, but thats the base I think.

Hope that helps!

------

I ll post an example I did last month while doing a cg workshop with Paul Mcwilliams, I had done this in softimage, but I wanted to learn maya, so that workshop was awesome for it!

this are buildings modeled in softimage, maya and zbrush using the photometry technique I ve explained above.










EDIT:

By the way I wonder if any of you guys have tried to use nuke 7 new modeling features to build a photogrammetric digital set. In case you have that ll be cool to know your thoughts on this!

Cheers

-Manuel

Last edited by manuqc : 01 January 2013 at 05:23 PM.
 
Old 02 February 2013   #18
Very Helpful!

Thank you! You've been very helpful...sorry for the delay. I'm not cool enough to offer kudos to your rep on the forum yet (apparently) but I would if I could. Is the workshop recording available for purchase? I had assumed the point cloud was generated through another means (scanning equipment of some sort). It's interesting to know that the points are generated directly from the photos. Concerning a building I'm assuming the 3 photos would have to be at different locations, and one would have to know a dimension of some sort, or is that picked up by the software / info in the photo depending on the camera and the 3 photos are simply to have the best accuracy?

Thanks again... your knowledge is priceless.
 
Old 02 February 2013   #19
We've used Stereo Aerial Photography to create about 150-200 acres of populated and developed area. The airplane makes a pass with multiple cameras and creates the data (XYZ) needed for the entire terrain, structures, treelines, lakes, etc as well as capturing the photos. It's a pretty slick technique. We've only done it once and it was probably about 8 years ago.

I'm don't know much about the technology, but I am sure it's bad ass government recon shit that has made its way down to the high end consumer level 20 years later.

We kicked the snot out of the opposing client's claim (I work in litigation graphics) and had a super happy client in the end

Last edited by XLNT-3d : 02 February 2013 at 03:04 PM.
 
Old 02 February 2013   #20
Thread automatically closed

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.
 
Thread Closed share thread



Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
CGSociety
Society of Digital Artists
www.cgsociety.org

Powered by vBulletin
Copyright 2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump
Miscellaneous

All times are GMT. The time now is 06:47 PM.


Powered by vBulletin
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.