Recovery and refinement of motion and structure from perspectively projected optical flow

Abstract

Determination of the 3D motion and structure of an object from its perspectively projected 2D image is one of the most important problems in computer vision. Previous works on shape and motion from optical flow have been unsatisfactory because they require solving nonlinear simultaneous equations using iterative search. In the present paper, we propose a method for the recovery of motion and structure from perspectively projected optical flow of feature points which move rigidly. In order to simplify the computation and obtain clear perspectives, we adopt a spherical image screen and arrange equations in linear form. As a result our method does not require the solution of nonlinear simultaneous equations, but only requires the solution of linear ones. Furthermore, we propose a method for refining the solution of our linear method using a nonlinear iterative search. Simulation results of our method are also presented.