I am doing a camera projection (from a picture onto a 3D object). In the picture I am projecting from, there are some hidden parts of the object due to the geometry of the object itself, and I would like to avoid Blender to "invent" those hidden parts of the texture by repeating other visible parts so I can know accurately and quickly which parts are still needed to be photographed in order to complete the full texture of the object.
My idea is to bake the projection over a green chroma texture but I do not want the UV project modifier to complete the texture regarding those hidden sections of the 3D piece (which is currently doing).
Does anybody know how to avoid that?
EDIT: Just discovered the image is being projected onto every voxel colliding with each ray from the camera, which means it is being projected onto backfaces where (in my opinion) it should not. Researching deeper in this, I just found a topic which describes my problem and potential solution:
It suggests to use object's negative Z coordinates to "blank" those parts of the texture and therefore avoid printing backfaces. Is there any other solution? Tell me if you consider to remove or to solve this thread or mark it as duplicated in case there is no other possible solution.
Thank you again for your feedback. Best regards, maije