I have a pretty complicated client requirement about generating a 3D model from 2D images, and although I have tried to find similar solutions I haven’t quite found exactly the answer to what I’m looking for, or at least I don’t seem to be going in the right direction.
The requirement is for a mobile phone user to be able to take multiple photos of themselves at various angles using their front camera, including the back of the head (I understand the user will require some form of assistance on the correct positioning, etc.). From this a 3D model needs to be generated of the complete head, including the back (not just the front, which I have found many articles on).
I am initially thinking about using OpenCV or a related framework to help with constructing the images and generating landmarks around the head, including the back.
Unfortunately, I am not able to use any algorithms that estimate a part of the head from previous trainings (I have found various research papers on this area on Arxiv and PapersWithCode, but sadly I can’t base the implementation on any of these); the requirement is for the head to be reconstructed based on the images taken by the mobile app.
Does anyone know of a suitable path I can follow to be able to implement this solution? Any advice on theory, papers and even code examples would be appreciated.
Source: Windows Questions C++