Hi,
new here.
I am experimenting with 2.5D / projections recently - and not sure where to ask -
so (after having been following Dr.Sassi ´s projection series for photographers,
and/or other Cineversity tutorials, for a while)-
I thought I´d try to ask here ..
I am trying to ´reconstruct´ a 3D model from an image
(just to add some ´2.5D parallax treatment - a bit of a depth/motion - to a 2D image)
Basically, the general workflow that I´ve learnt so far, already works pretty nicely for other projects of mine (architecture, matte-paintings, etc.)
But - this time - it´s different.
It´s an illustration/concept art image, not a photo -
(single image only / illustration of a fantasy ´tank-machine´, depicted in 3-point perspective)
So - unlike with a photo -
none of those little real-world technical ´clues´ and/or information, that real photography/camera capture can provide - to be ´extracted´ - are available here ..
And/Or - those that are available - are actually not very helpful, really .. (almost no right angle, partial or missing symmetry, loose perspective, etc.) ...
So - any kind of ´reverse engeneering´ - that otherwise works for other cases - seem to either fail - or be uneffective here, for me, so far.
Whatever I may try - it seems to end up with a pretty distorted model/projection setup, in the end.
Just thought there must (?) be some better solution, maybe ..
So my question:
What is/could be the best approach in this case - if any?
In general?
Maybe I am still missing something?
Just curious ..
Thanks,
alena h.
Image preview:
https://www.dropbox.com/s/s5wkqvyfvhnimgi/mk-final.jpg?dl=0