mapping photos to mesh

[jm328]

Hi,

I am trying to map several photos onto a mesh model. 

The UVW map modifier works well. I can position one photo, but it applies it to the entire model. I'd like to have it apply just to the faces that the photo covers. I found this is possible in per pixel mapping through camera map modifier, but had trouble aligning the photos.  

Here's the model created from scan data (I still need to repair some holes) http://www.thetemporaryarchive.org/example_copy.html

How would you tackle mapping this model with photos? I have 22 totem poles to do. Should I unwrap the meshes? That seems way too time consuming. 

Thanks for any suggestions!

[grue1970]

i can't see the mesh you have as the link shows up as a black screen with a title across the top.

[jm328]

Sorry about that. It displays using WebGL. It is only viewable in Chrome,
Firefox, or Safari (after enabling WebGL). It doesn't work in IE. It takes
a few moments or minutes to load. Then a popup will appear that says "hit
me". Click that and it should show.

Let me know if it still doesn't work.

Thanks.

[jm328]

One note about viewing the model. 

It displays using WebGL. It is only viewable in Chrome, Firefox, or Safari (after a few steps to enable WebGL). Unfortunately, it doesn't work in IE. It takes a few moments to load. Then a popup will appear that says "hit me". Click that and it should display.

You should have mouse orbit controls. Pan, zoom, 3d orbit.

[grue1970]

ah, i wasn't patient enough. yikes, i see what you have there. hm....  are your photos of that same object (colors and such)?  could they be stitched together vertically and then just apply it that way?

[grue1970]

here's a screen shot.  so what did you do to scan that?  you know, autodesk has some type of image-to-3d online beta (i think it's beta) called recap photo.  https://recap360.autodesk.com/signin.aspx#home

it looks like this might be the perfect reason it was created 🙂  and in their examples it makes the object with material that matches the photos.

[jm328]

Thanks for the reply. I'm curious to see what the results of the Recap photo modeler would be. I don't think I have enough shots from different angles to make it work though. All my photos are from five scan positions. They are taken from the same nodal point of the scanner. 

I have vertex colors from the pointcloud, so I could burn a texture with that. It's not as high resolution as re-mapping photos to the mesh though. I might try your suggestion to combine the photos into a single orthographic image for each side.  Thanks for the suggestions.