On the Two-View Geometry of Unsynchronized...

On the Two-View Geometry of Unsynchronized Cameras

https://openaccess.thecvf.com/content_cvpr_2017/papers/Albl_On_the_Two-View_CVPR_2017_paper.pdf

Abstract
We present new methods for simultaneously estimating camera geometry and time shift from video sequences from multiple unsynchronized cameras. Algorithms for simultaneous computation of a fundamental matrix or a homography with unknown time shift between images are developed. Our methods use minimal correspondence sets (eight for fundamental matrix and four and a half for homography) and therefore are suitable for robust estimation using RANSAC. Furthermore, we present an iterative algorithm that extends the applicability on sequences which are significantly unsynchronized, finding the correct time shift up to several seconds. We evaluated the methods on synthetic and wide range of real world datasets and the results show a broad applicability to the problem of camera synchronization

Authors and associations to note:

Jan Heller
CTO of Magik-Eye

Researcher in the field of artificial intelligence with an expertise in computer vision, and robotics camera calibration.

Tomas Pajdla
Scientific Advisor at Magik-Eye
Assistant Professor, Czech Technical University, Department of Cybernetics
Distinguished Researcher for Center of Machine Perception
Co-Founder, Neovision

Andrew Fitzgibbon

HoloLens, Microsoft
Cambridge


Figured this deserved its own thread for discussion.