Journal of System Simulation
Abstract
Abstract: Aiming at dealing with the problem of natural image matting, this study focused on a closed-form solution to image pairs matting. For given two images and their corresponding trimaps, the affinity matrix between the two images is calculated. The matting Laplacian which characters an image to co-matting Laplacian which characters a pair of images is generalized. The alpha mattes of image pairs were obtained by optimizing the energy function which derived from co-matting Laplacian. To validate the approach, the experiments on several datasets are conducted. The results demonstrate that the method can achieve higher quality and is suitable for matting of a sequence of images.
Recommended Citation
Yi, Wei; Liu, Shang; Qiu, Jiahe; Zhang, Guijuan; Zhu, Dengming; and Shen, Yanfei
(2019)
"Co-matting of Natural Image Pairs,"
Journal of System Simulation: Vol. 30:
Iss.
11, Article 16.
DOI: 10.16182/j.issn1004731x.joss.201811016
Available at:
https://dc-china-simulation.researchcommons.org/journal/vol30/iss11/16
First Page
4180
Revised Date
2018-06-10
DOI Link
https://doi.org/10.16182/j.issn1004731x.joss.201811016
Last Page
4186
CLC
TP391.9
Recommended Citation
Wei Yi, Shang Liu, Qiu Jiahe, Zhang Guijuan, Zhu Dengming, Shen Yanfei. Co-matting of Natural Image Pairs[J]. Journal of System Simulation, 2018, 30(11): 4180-4186.
DOI
10.16182/j.issn1004731x.joss.201811016
Included in
Artificial Intelligence and Robotics Commons, Computer Engineering Commons, Numerical Analysis and Scientific Computing Commons, Operations Research, Systems Engineering and Industrial Engineering Commons, Systems Science Commons