A robust linear least-squares estimation of camera exterior orientation using multiple geometric features

For photogrammetric applications, solutions to camera exterior orientation problem can be classified into linear (direct) and non-linear. Direct solut ions are important because of their computational efficiency. Existing line ar solutions suffer from lack of robustness and accuracy partially due to t he fact that the majority of the methods utilize only one type of geometric entity and their frameworks do not allow simultaneous use of different typ es of features. Furthermore, the orthonormality constraints are weakly enfo rced or not enforced at all. We have developed a new analytic linear least- squares framework for determining camera exterior orientation from the simu ltaneous use of multiple types of geometric features. The technique utilize s 2D/3D correspondences between points, lines, and ellipse-circle pairs. Th e redundancy provided by different geometric features improves the robustne ss and accuracy of the least-squares solution. A novel way of approximately imposing orthonormality constraints on the sought rotation matrix within t he linear framework is presented. Results from experimental evaluation of t he new technique using both synthetic data and real images reveal its impro ved robustness and accuracy over existing direct methods. (C) 2000 Elsevier Science B.V. All rights reserved.