Depth-based image completion for View Synthesis
J. Gautier, O. Le Meur, C. Guillemot
3DTV 2011
Main Idea
This paper describes a depth-based inpainting algorithm which efficiently handles disocclusion occurring on virtual viewpoints rendering. A single reference view and a set of depth maps are used in the proposed approach. The method not only deals with small disocclusion filling related to small camera baseline, but also manages to fill in the synthesized views by coherent tensor-based color and geometry structure propagation. The depth is used to drive the filling order, while enforcing the structure diffusion from similar candidate-patches. By acting on patch prioritization, selection and combination, the completion of distant synthesized views allows a consistent and realistic rendering.Videos
[WEBM-VP8] [M4V-MP4]
Original figures of the paper: Ballet
under construction
Figure a: V5->4 after warping and background antighosting. b: V5->2 after warping and background antighosting.
Figure c: V5->4 inpainted with Criminisi's method. d: V5->2 inpainted with Criminisi's method.
Figure e: V5->4 inpainted with Daribo's method.f: V5->2 inpainted with Daribo's method.
Figure g: V5->4 inpainted with the proposed method.g: V5->2 inpainted with the proposed method.
Additionnal figures: Breakdancers
Figure a: V5->4 after warping and background antighosting. b: V5->2 after warping and background antighosting.
Figure c: V5->4 inpainted with Criminisi's method. d: V5->2 inpainted with Criminisi's method.
Figure e: V5->4 inpainted with Daribo's method.f: V5->2 inpainted with Daribo's method.
Figure g: V5->4 inpainted with the proposed method.g: V5->2 inpainted with the proposed method.