Self-calibrating cameras using semidefinite programming

Abstract

Novel methods are proposed for self-calibrating a pure-rotating camera using semidefinite programming (SDP). Key to the approach is the use of the positive-definiteness requirement for the dual image of the absolute conic (DIAC). The problem is couched within a convex optimization framework and convergence to the global optimum is guaranteed. Experiments on various data sets indicate that the proposed algorithms more reliably deliver accurate and meaningful results. This work points the way to an alternative and more general approach to self-calibration using the advantageous properties of SDP. Algorithms are also discussed for cameras undergoing general motion.