|
|
- 2016
多类变分优化的自然图像分割方法
|
Abstract:
针对自然图像中内容的多样性、复杂性以及随机性，若采用区域内部恒定聚类中心假设的CV（Chan-Vese）模型以及多类水平集模型，则难以有效刻画具有非线性、连续性变化的自然图像内容。该文通过对区域内部自由度调控的多变量学生-t概率密度分布描述，提出了多类非线性变分活动轮廓模型，它打破了区域内部恒定密度的约束。由于多类非线性变分活动轮廓模型缺乏区域外力，容易分割出离散、零碎的噪声区域，通过引入测地线区域外力约束项，能有效分割出区域间的光滑边界。针对多类变分模型的最小化问题是NP难问题，提出对多类变分活动轮廓模型进行离散化表达，然后构建对应的多层图割模型，并利用最大流/最小割优化方式快速求得全局近似最优解。实验表明，该文提出的分割方法能够准确地分割出多类非同质目标区域，且区域之间的边界光滑，视觉效果好。
| [1] | LANKTON S, TANNENBAUM A. Localizing region-based active contours[J]. IEEE Transactions on Image Processing, 2008, 17(11):2029-2039. |
| [2] | 李伟斌, 高二, 宋松和. 一种全局最小化的图像分割方法[J]. 电子与信息学报, 2013, 35(4):791-796. LI Wei-bin, GAO Er, SONG Song-he. A global minimization method for image segmentation[J]. Journal of Electronics & Information Technology, 2013, 35(4):791-796. |
| [3] | 郑锦, 仙树, 李波. 基于形状约束和局部演化的二值水平集运动目标分割[J]. 电子与信息学报, 2013, 35(5):1037-1043. ZHENG Jin, XIAN Shu, LI Bo. Moving object segment-ation using binary level set based on shape constraint and local evolution[J]. Journal of Electronics & Information Technology, 2013, 35(5):1037-1043. |
| [4] | LEINER B J, RAMIREZ B E, VALLEJO E. Comparative study of variational and level set approaches for shape extraction in cardiac CT images[C]//International Seminar on Medical Information Processing and Analysis.[S.l.]:SPIE, 2014. |
| [5] | ZHANG T, FREEDMAN D. Tracking objects using density matching and shape priors[C]//IEEE International Conference on Computer V ision.[S.l.]:IEEE, 2004:1950-1954. |
| [6] | CASELLES V, CATTE F, COLL T, et al. A geometric model for active contours in image processing[J]. Numerische Mathematik, 1993, 66:1-31. |
| [7] | MUMFORD D, SHAH J. Optimal approximations by piecewise smooth functions and associated variational problems[J]. Communications on Pure and Applied Mathematics, 1989, 42:577-685. |
| [8] | 全刚, 孙即祥. 基于活动轮廓的图像分割方法研究[D].长沙:国防科技大学, 2010. QUAN gang, SUN Ji-Xiang. Image segmentation method based on active contour[D]. Changsha:National University of Defense Technology, 2010. |
| [9] | TAO W B, CHANG F, LIU L M, et al. Interactively multiphase image segmentation based on variational formulation and graph cuts[J]. Pattern Recognition, 2010, 43(10):3208-3218. |
| [10] | KASS M, WITKIN A, TERZOPOULOS D. Snakes:Active contour models[J]. International Journal of Computer Vision, 1988, 1:321-331. |
| [11] | YANG Yong, HAN Shou-dong, WANG Tian-jiang, et al. Multilayer graph cuts based unsupervised color-texture image segmentation using multivariate mixed student's t-distribution and regional credibility merging[J]. Pattern Recognition, 2013, 46(4):1101-1124. |
| [12] | ISHIKAWA H. Exact optimization for markov random fields with convex priors[J]. IEEE Transactions on Transactions on Pattern Analysis and Machine Intelligence, 2003, 25(10):1333-1336. |
| [13] | BAE E, TAI Xue-cheng. Graph cuts for the multiphase Mumford-Shah model using piecewise constant level set methods[EB/OL].[2014-01-20]. http://www.doc88.com/p-303734285052.html. |
| [14] | BOYKOV Y, KOLMOGOROV V. Computing geodesics and minimal surfaces via graph cuts[C]//IEEE International Conference on Computer V ision. Nice, France:IEEE, 2003:1-8. |
| [15] | BOYKOV Y, VEKSLER O. ZABIH R. Fast approximate energy minimization via graph cuts[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2001, 23(11):1222-1239. |
| [16] | FULKERSON D, FOLD L. Flow in networks[R].[S.l.]:Princeton University Press, 1962. |
| [17] | UNNIKRISHNAN R, PANTOFARU C. HEBRET M. Toward objective evaluation of image segmentation algorithms[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 29(6):929-944. |
| [18] | VESE L, CHAN T. A multiphase level set framework for image segmentation using the mumford and shah model[J]. International Journal of Computer Vision, 2002, 50(3):271-293. |
| [19] | CHAN T, VESE L. Active contours without edges[J]. IEEE Transactions on Image Processing, 2001, 10(2):266-277. |
| [20] | BYEONG R L, BEN A H, HAMID K. An active contour model for image segmentation:a variational perspective[C]//IEEE International Conference on Acoustics, Speech, & Signal Processing.[S.l.]:IEEE, 2002, 2:1585-1588. |
