"GrabCut": interactive foreground extraction using iterated graph cuts
ACM SIGGRAPH 2004 Papers
High performance imaging using large camera arrays
ACM SIGGRAPH 2005 Papers
Natural video matting using camera arrays
ACM SIGGRAPH 2006 Papers
Graph Cuts and Efficient N-D Image Segmentation
International Journal of Computer Vision
Reconstructing Occluded Surfaces Using Synthetic Apertures: Stereo, Focus and Robust Measures
CVPR '06 Proceedings of the 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition - Volume 2
CVPR '11 Proceedings of the 2011 IEEE Conference on Computer Vision and Pattern Recognition
Hi-index | 0.00 |
This paper proposes a novel method to see-through occlusion and automatically focuses on object with high-level imaging quality using camera array. Even with the amazing perspective identity, synthetic aperture imaging still suffers from blurs and disturbance caused by occlusion. The novelties of the approach include: (1) Rather than using the direct observed images to achieve synthetic aperture image, this paper raises the idea to synthesize edge image, which synthetic binary images after edge detection. (2) Based on the special data identity of camera array, this paper proposes an "Auto-Cut" segmentation idea, which could upgrade interactive cut method, such as GrowCut, GrabCut and Graph Cut, to a totally automatic method. (3) This paper proposes an automatically selecting the focal depth method which could yield a convincing estimation even under serious occluded situation. The feasibility of our approach is experimentally demonstrated. A multi-view images based improved synthetic aperture imaging system has been set up, and experimental results with qualitative and quantitative analysis demonstrate that the method can improve imaging quality and resist occlusion in challenge scene.