Linear Distributed Localization of Omni-Directional Camera Networks

Localization, estimating the positions and orientations of a set of camera nodes, is a critical first step in camera network applications such as geometric estimation and scene reconstruction. We propose a distributed algorithm for camera network localization between multiple omni-directional cameras with sparse overlapping view structure. This procedure involves a distributed variation of the bundle adjustment algorithm with a linear error metric tailored for omnidirectional cameras. The linear structure of the system allows us to show convergence up to a scale factor of the iterative algorithm through an SVD-based approach and provide bounds on convergence rates. Finally, we show simulated and experimental results showing on-par performance with modern centralized techniques.