In order to deal with the color overlap problem in matting, a fast random projection method is proposed to complement the color information. First, the raw texture matrix is obtained through dense abstraction from color image. The random projection is performed and the best three texture channels are chosen by the foreground and background overlap factors. Combining the texture image, the new cost function takes into account texture, color, and spatial information. Second, the filtering process is carried out to the sample selection cost, including the effect of the local and nonlocal neighbors. Finally, the relationship between iterative filter and global energy smooth is proven, and the post filter formula is obtained. Experiments show that the cost filtered matting with random texture features produces both visually and quantitatively better results when the color distributions of the foreground and background are similar.
Levin A, Lischinski D, and Weiss Y. A closed form solution to natural image matting[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, 30(2): 61-68.
[2]
Shi Y, Au O C, Pang J, et al.. Color clustering matting[C]. IEEE International Conference on Multimedia and Expo, California, USA, 2013, 7: 1-6.
[3]
Wang J and Cohen M F. Optimized color sampling for robust matting[C]. IEEE Conference on Computer Vision and Pattern Recognition, Minneapolis, USA, 2007, 6: 281-288.
[4]
He B,Wang G J, and Zhang C. Iterative transductive learning for automatic image segmentation and?matting?with RGB-D data[J]. Journal of Visual Communication and Image Representation, 2014, 25(5): 1031-1043.
[5]
Shahrian E, Rajan D, Price B, et al.. Improving image matting using comprehensive sampling sets[C]. Conference on Computer Vision and Pattern Recognition, Oregon, Portland, USA, 2013, 6: 636-643.
[6]
Shahrian E and Rajan D. Weighted color and texture sample selection for image matting[C]. IEEE Conference on Computer Vision and Pattern Recognition, Rhode Island, USA, 2012, 6: 718-725.
[7]
Rhemann C, Rother C, and Gelautz M. Improving color modeling for alpha matting[C]. The British Machine Vision Conference, Leeds, UK, 2008, 9: 1155-1164.
[8]
He K, Rhemann C, Rother C, et al.. A global sampling method for alpha matting[C]. IEEE Conference on Computer Vision and Pattern Recognition, Colorado, USA, 2011, 6: 2049-2056.
[9]
Gastal E S and Oliveira M M. Shared sampling for real‐time alpha matting[J]. Eurographics, 2010, 29(2): 575-584.
[10]
Jubin J, Deepu R, and Hisham C. Sparse codes as alpha mattes[C]. The British Machine Vision Conference, Nottingham, England, 2014, 9: 1-11.
[11]
Varma M and Zisserman A. A statistical approach to material classification using image patches[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(11): 2032-2047.
[12]
Liu L and Paul W. Texture classification from random features[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(3): 574-586.
[13]
Hosini A, Bleyer M, Rother C, et al.. Fast cost-volume filtering for visual correspondence and beyond[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(2): 504-511.
[14]
Dasgupta S and Gupta A. An elementary proof of a theorem of Johnson and Lindenstrauss[J]. Random Structures and Algorithms, 2003, 22(1): 60-65.
[15]
Sural S, Qian G, and Pramanik S. Segmentation and histogram generation using the HSV color space for image retrieval[C]. International Conference on Image Processing, New York, USA, 2002, 2: 589-592.
[16]
Marius M and David GL. Fast approximate nearest neighbors with automatic algorithm configuration[C]. International Conference on Computer Vision Theory and Applications, Lisbon, Portugal, 2009, 2: 331-340.
[17]
Rhemann C, Rother C, Wang J, et al..?A perceptually motivated online benchmark for image matting[OL]. http://www.alphamatting.com. 2009, 5.
