Depth Driven Photometric and Geometric Image Registration for Real-Time Stereo Systems

Abstract

This paper presents a novel depth driven approach for a highly accurate joint photometric and geometric image alignment. Thereby, the registration problem is expressed in terms of a consistent, elegant and efficient energy formulation. Moreover, we propose a real-time capable alternating iterative optimization scheme to solve for a state of minimal energy. Since our energy formulation is based on pixel wise color similarity our registration procedure directly improves the performance of disparity estimation and the visual quality of multi-view view synthesis.