Superpixel-based depth map inpainting for RGB-D view synthesis
Authors:
- 1 September 2015
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 4332-4336
Abstract
In this paper we propose an approach to inpaint holes in depth maps that appear when synthesizing virtual views from a RGB-D scenes. Based on a superpixel oversegmentation of both the original and synthesized views, the proposed approach efficiently deals with many occlusion situations where most of previous approaches fail. The use of superpixels makes the algorithm more robust to inaccurate depth maps, while giving an efficient way to model the image. Extensive comparisons to relevant state-of-the-art methods show that our approach outperforms qualitatively and quantitavely these existing approaches.Keywords
This publication has 10 references indexed in Scilit:
- Eikonal-based region growing for efficient clusteringImage and Vision Computing, 2014
- High-Resolution Stereo Datasets with Subpixel-Accurate Ground Truth2014
- Inter-view consistent hole filling in view extrapolation for multi-view image generation2014
- Depth-Based Disocclusion Filling for Virtual View Synthesis2012
- A new multidirectional extrapolation hole-filling method for Depth-Image-Based Rendering2011
- Depth-based image completion for view synthesis2011
- Depth Image-Based Rendering With Advanced Texture Synthesis for 3-D VideoIEEE Transactions on Multimedia, 2011
- Depth-aided image inpainting for novel view synthesis2010
- O(N) implementation of the fast marching algorithmJournal of Computational Physics, 2006
- Region Filling and Object Removal by Exemplar-Based Image InpaintingIEEE Transactions on Image Processing, 2004