Analysis of Symmetric Panorama Acquisition

Abstract

This paper reports about basic algebraic relations between parameters and an error analysis for symmetric panorama acquisition. Symmetric panoramas are of importance in computer vision, computer graphics, and stereoscopic imaging and display. Advantages of symmetric panoramas include the possible reuse of stereo matching algorithms previously developed for 3D reconstruction, and possible applications in stereoscopic visualization. This paper formulates and studies problems which have not yet been approached previously: algebraic relations between application-specific parameters and imaging parameters, and how errors (incurred from measurements during imaging) are propagating and impacting the quality of resultant images. Without dealing with such problems, we are not able to answer the more difficult questions regarding the design and/or the capability assessment of symmetric panorama acquisition systems. The paper first summarizes the acquisition geometry followed by in-depth studies of algebraic parameter relations and error analyses.