Absolute Pose Estimation of Central Cameras Using Planar Regions

A novel method is proposed for the absolute pose estimation of a central 2D camera with respect to 3D depth data without use any dedicated calibration pattern or explicit point correspondences. The has no specific assumption about source: plain information expected from sensing device and used capture images. Both perspective omnidirectional cameras are handled within single generic model. Pose formulated as 2D-3D nonlinear shape registration task which solved correspondences complex similarity metrics. It relies on set corresponding planar regions, parameters obtained by solving an overdetermined system equations. efficiency robustness were confirmed both large scale synthetic real acquired various types sensors.