Structure from Motion with Known Camera Positions

Abstract

The wide availability of GPS sensors is changing the landscape in the applications of structure from motion techniques for localization. In this paper, we study the problem of estimating camera orientations from multiple views, given the positions of the viewpoints in a world coordinate system and a set of point correspondences across the views. Given three or more views, the above problem has a finite number of solutions for three or more point correspondences. Given six or more views, the problem has a finite number of solutions for just two or more points. In the three-view case, we show the necessary and sufficient conditions for the three essential matrices to be consistent with a set of known baselines. We also introduce a method to recover the absolute orientations of three views in world coordinates from their essential matrices. To refine these estimates we perform a least-squares minimization on the group cross product SO(3) × SO(3) × SO(3). We report experiments on synthetic data and on data from the ICCV2005 Computer Vision Contest. ©2006 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE.

Multi-view structure computation without explicitly estimating motion

“…the coordinates of the 3D point clouds, and we are done. If camera motion (location and orientations) is further needed, one can simply solve it from these solved 3D points using PnP or camera pose methods [7,8,5]. …”

“…The Corridor sequence contains 206 3D-points 5 extracted from 11 views. To construct the rigid graph used in our method, we collect all available edges.…”

“…This removes the slack caused by the inherent uncertainty of cam- 5 More precisely, image point tracks. era motion estimation.…”

“…Researches on this topic are still active now (e.g. [11][20] [5][18] [14]). This paper proposes a radically new approach to Sfm, which bypasses the motion-estimation step, and goes directly to the 3D structure computation stage.…”

Orientation and Pose recovery from Spherical Panoramas

“…Carceroni et al present a method that relies on some properties of SO(3) [3]. It is a feature-based procedure that uses a GPS to recover camera orientation from known positions of multiple cameras.…”