Uncalibrated vision for 3-D underwater applications

We present an uncalibrated vision system recovering 3D reconstructions of underwater scenes using no camera calibration at all. The only information available is a set of point correspondences plus some a-priori knowledge on the scene structure. The system includes three main components: a robust tracker, which targets optimal features for video tracking, and implements a completely automatic mechanism for rejecting unstable image features; a module computing a projective reconstruction of the scene from the set of features declared reliable by the tracker; and a module recovering the target Euclidean reconstruction from the projective one, employing a-priori knowledge of the distances between five identifiable scene points. We sketch the design of each component and report experiments with our working implementation on real data sets, collected by cameras immersed in our laboratory tank.