Robust upright adjustment of 360 spherical panoramas

Abstract

With the recent advent of 360 cameras, spherical panorama images are becoming more popular and widely available. In a spherical panorama, alignment of the scene orientation to the image axes is important for providing comfortable and pleasant viewing experiences using VR headsets and traditional displays. This paper presents an automatic method for upright adjustment of 360 spherical panorama images without any prior information, such as depths and gyro sensor data. We take the Atlanta world assumption and use the horizontal and vertical lines in the scene to formulate a cost function for upright adjustment. In addition to fast optimization of the cost function, our method includes outlier handling to improve the robustness and accuracy of upright adjustment. Our method produces visually pleasing results for a variety of real-world spherical panoramas in less than a second, and the accuracy is verified using ground-truth data. ? 2017, Springer-Verlag Berlin Heidelberg.