3D Spherical Panoramic Epipolar Line Based on Essential Matrix

Abstract

A 3D spherical panoramic epipolar line based on essential matrix is proposed in this study to solve the search range to improve calculation efficiency and reliability of the matching. First, this method uses the essential matrix of computer vision principles to establish the epipolar geometry model of spherical panoramic image. Second, it derives the epipolar line mathematical equations and characteristics of spherical panoramas. Third, it calculates statistics on the distribution law of epipolar images. Experimental results show that the feasibility of the method proposed in this study. The proposed method effectively verified the trajectory of the spherical panoramic epipolar line, finished distribution of epipolar line sequence trajectory and decreased the search range of the corresponding points. The findings help lay a solid foundation for the matching and measurement of the panoramic measurement model and the generation of depth maps.