Chirality in Computer Vision

  • Rainer Sinn (Universität Leipzig)
E1 05 (Leibniz-Saal)


A fundamental question in computer vision is whether a set of point pairs is the image of a scene that lies \textbf{in front} of cameras. Such a scene and the cameras together are known as a chiral reconstruction of the point pairs. We discuss how this constraint “to be in front of" can be expressed in terms of polynomial inequalities in real projective space. In particular, we focus on the case of two cameras which has connections to classical (real) geometry of cubic hypersurfaces.