Fast Matting Using Large Kernel Matting Laplacian Matrices

被引：113

作者：

He, Kaiming ^{[1
]}

Sun, Jian ^{[2
]}

Tang, Xiaoou ^{[1
,3
]}

机构：

[1] Chinese Univ Hong Kong, Dept Informat Engn, Hong Kong, Hong Kong, Peoples R China

[2] Microsoft Res Asia, Beijing, Peoples R China

[3] Chinese Acad Sci, Shenzhen Inst AdvTech, Beijing 100864, Peoples R China

来源：

2010 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR) | 2010年

关键词：

D O I：

10.1109/CVPR.2010.5539896

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Image matting is of great importance in both computer vision and graphics applications. Most existing state-of-the-art techniques rely on large sparse matrices such as the matting Laplacian [12]. However, solving these linear systems is often time-consuming, which is unfavored for the user interaction. In this paper, we propose a fast method for high quality matting. We first derive an efficient algorithm to solve a large kernel matting Laplacian. A large kernel propagates information more quickly and may improve the matte quality. To further reduce running time, we also use adaptive kernel sizes by a KD-tree trimap segmentation technique. A variety of experiments show that our algorithm provides high quality results and is 5 to 20 times faster than previous methods.

引用

页码：2165 / 2172

页数：8