View-consistent 4D Light Field Depth Estimation

09/09/2020
by   Numair Khan, et al.
14

We propose a method to compute depth maps for every sub-aperture image in a light field in a view consistent way. Previous light field depth estimation methods typically estimate a depth map only for the central sub-aperture view, and struggle with view consistent estimation. Our method precisely defines depth edges via EPIs, then we diffuse these edges spatially within the central view. These depth estimates are then propagated to all other views in an occlusion-aware way. Finally, disoccluded regions are completed by diffusion in EPI space. Our method runs efficiently with respect to both other classical and deep learning-based approaches, and achieves competitive quantitative metrics and qualitative performance on both synthetic and real-world light fields

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset