Close-up View synthesis by Interpolating Optical Flow
This addresses the challenge of virtual navigation by enabling pseudo-3D projection without depth information, offering a practical solution for applications like street view systems.
The paper tackles the problem of synthesizing close-up virtual views without depth sensors by using optical flow to create parallax effects, achieving clear and visually faithful magnified results that overcome distortion and blur issues in systems like Google Street View.
The virtual viewpoint is perceived as a new technique in virtual navigation, as yet not supported due to the lack of depth information and obscure camera parameters. In this paper, a method for achieving close-up virtual view is proposed and it only uses optical flow to build parallax effects to realize pseudo 3D projection without using depth sensor. We develop a bidirectional optical flow method to obtain any virtual viewpoint by proportional interpolation of optical flow. Moreover, with the ingenious application of the optical-flow-value, we achieve clear and visual-fidelity magnified results through lens stretching in any corner, which overcomes the visual distortion and image blur through viewpoint magnification and transition in Google Street View system.