Some Improvements on Two Autocalibration Algorithms Based on the Fundamental Matrix

Autocalibration algorithms based on the fundamental matrix must solve the problem of finding the global minimum of a cost function which has many local minima. We describe a new method of achieving this goal, which uses a stochastic optimization approach taken from the field of evolutionary algorithms. In theory, approaches that use the fundamental matrix for autocalibration are inferior to those based on a projective reconstruction. We argue that in practice if we use this new stochastic optimization approach this is not true. When autocalibrating focal length and aspect ratio both methods achieve comparable results. We demonstrate this experimentally using published image sequences for which the ground truth is known.