It´s obvious: the main problem of 3D astrophysics is to find contraints on the structure along the line of sight. Of course, but does this problem get sufficient systematic research attention? Some of the constraints will come from the observations, others from theoretical assumptions on the objects that will have to complement the limited constraints from data. Assessing the quality of the available constraints is non-trivial and should get a lot of attention and research to improve it, since the constraints are at the heart of the problem of 3D reconstruction. How can we improve the constraints? Administrators would reply: Let´s set up a commission. Astronomers say: Let´s find a nice beach resort to have a conference so we can talk it over. Any suggestions?…for the beach, of course. Please read on, we need your feedback.
3D reconstruction from single images is a big topic in Computer Graphics. Of course, usually it deals with objects from our everyday environment, such as buildings, faces, furniture or other common objects. A first step in such a reconstruction commonly is an object recognition algorithm that identifies the type of object that one is dealing with. Then from the known general structure a more detailed 3D structure can be infered after deriving its orientation. Can we do this in astrophysics? Continue reading