Setting up multiple camera-projected images

Become a member of the CGSociety

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

THREAD CLOSED
 
Thread Tools Display Modes
  07 July 2013
Setting up multiple camera-projected images

Hello everyone,

I wanted to ask if there's an easy way to set up multiple camera-projected reference images (and switch between them). I know how to do this manually (create image plane, align to camera view, project the image). But I've got several .fbx files here with 60-110 cameras each and a reference image for everyone of them. So if there's any method or script that would save me the trouble of doing this by hand, I'd be very glad. I'm using Max 2014 64bit by the way.

Any help is appreciated.
__________________
 
  07 July 2013
Here's an easier method for managing projections in max:

http://www.projectgemini.net/CameraMapGemini/

but that won't help with your problem of setting it up.
 
  07 July 2013
Thanks for your reply! I agree, Gemini might help with managing multiple projections, but not with the setup itself. Also it doesn't seem to be maintained anymore?
I was hoping there was a cool script or plugin for the whole setup process, but I guess not. Maybe parts of it could be scripted, like creating, naming and aligning one plane for each camera...
__________________
 
  07 July 2013
Have you got an image plane script? there's one here: http://scripts.breidt.net/

I'm interested in finding one for Vray.
 
  07 July 2013
Any idea how the images are blended?
For example, if you have images of the same area from different angles, how is the overlap handled?

David
__________________
Nothing to see here

www.dustinforensics.com
 
  07 July 2013
Use a composite map and blend them as you would any other map. I like to make alpha maps, or use vertex maps to control where I see one or the other.
__________________
--Pel
 
  07 July 2013
Or the camera map Gemini has an angle threshold in the map but I'm unsure how it works exactly. I didn't set up any masks though.
 
  07 July 2013
Wow! Thanks guys, especially MisterS, thats exactly what I've been looking for. I'll try it tomorrow!

None of the pictures are blended or would need to be. They are pictures from photographic 3D scans (using Autodesk 123D Catch) with lots of intended overlap. I need to set up image planes to check the generated model and correct mistakes. Some difficult areas I'll probably have to re-model by hand, but I want to be able to constantly re-check the model against the reference planes from many angles.

The situation is a bit like the one on the first page of the amazing WIP Thread "crossing the uncanny valley" (though not on the same level).
__________________
 
  07 July 2013
Glad it helped. So how do you rate 123d? I used it's older version Imagemodeler a bit but AD canned it and it had a nasty bug in the FBX export that they said was unlikely to be looked at.
 
  07 July 2013
I was quite impressed by the results I finally got out of 123D Catch, but it took some trial and error to get there.
After a while you get a feeling for what it needs and how to tweak your shooting: clear, recognizable backgrounds, consistent lighting (no flash!). Avoid parts of the scene moving or being out of focus (depth of field) to help it find markers even in the background.
It has the annoying tendency to assume dark areas are shadows and to read them as depth, so if you scan a flat board with a black and white grid, the black lines will come out slightly indented.
But even if it needs some fixing (and retopology) you get a very detailed mesh out of it. Both FBX and OBJ export work fine as far as I can tell.
I haven't tried any of the other apps of the 123D suite.
__________________
 
  07 July 2013
It might be worth your time to investigate Photosynth toolkit and Synthexport

http://www.visual-experiments.com/2...osynth-toolkit/
https://synthexport.codeplex.com/

Photosynth creates your sparse point cloud and solves for your camera positions, photosynth toolkit creates a dense point cloud from the sparse point cloud, then Synthexport creates a maxscript that automates bringing the cameras/mapping into 3ds Max.

Admittedly, it's all a bit obtuse, but if you can figure it out, you'll likely get better results than you could with 123D catch.
 
  07 July 2013
@ MisterS: Thanks again, the script works perfectly.

@ BradT: Thanks for pointing out Photosynth. Looks very powerful. For now, due to time constraints I will probably have to make do with the meshes I already got out of 123D. But hopefully I'll be able to check out Photosynth in the future. A side-by-side comparison using the same set of photos would be interesting.
__________________
 
  07 July 2013
So what is the advantage of the image plane script?

Fusion has those by default, but, honestly, I would just use a background image and see it only in the viewport. In Fusion I use them just to I get fast feedback in the OGL viewport. I might leave the camera linked plane there as I set the scene up, but then always render with alpha and comp over the base image.

Or are you using the plane for more than a background image... Like to project something with an alpha. Even so I just through any old plane (or more complex geometry as needed) out there and camera map it.

For basic camera mapping stuff not requiring lighting Fusion has pretty much replaced other true 3d apps for our workflow.
__________________
--Pel
 
  07 July 2013
You can model using multiple cameras/photographs without the need to keep switching out your background image.
 
  07 July 2013
The resulting dome of planes helps me visualize how the reference is distributed and where there might be gaps. Like MisterS I also thought that it's faster than switching your background. But now that you mention it, I could of course try to create an animated background out of all the images and an animated camera aligned to all the others in turn and cycle through them using the timeline. But then I couldn't show different camera views in different viewports (As far as I know all viewports are at the same frame of an animation).

The planes created by the image plane script perform just like ordinary ones, it just saves a little bit of time by automating the aligning/scaling step and gives you a few parametric options. Which is nice if you have to do this several hundred times. It would have been even better if the script could create planes for multiple cameras in one step. Maybe this would be the right time to enter the dangerous waters of scripting Somehow it should be possible to automatically run the script until it runs out of cameras in the scene.

Unfortunately I don't have access to Fusion here.
__________________
 
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 03:12 AM.


Powered by vBulletin
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.