基于全散度的变分CV模型及其分割算法

doi:10.11896/j.issn.1002-137X.2015.04.063

摘要/Abstract

摘要： CV模型在图像灰度不均匀或有噪声干扰时,易出现错分现象,因此将全散度引入变分CV模型,提出了基于全散度的变分CV模型及其迭代分割算法。分析基于欧氏距离所对应的变分CV模型分割算法存在的问题和不足,通过图示说明全散度相对于欧氏距离在距离计算与坐标系选择无关的优势,将其引入变分CV模型拟合偏差项,来提高图像灰度值与分割区域平均灰度偏差计算的鲁棒性。然后,采用欧拉-拉格朗日变分法获得全散度变分CV模型的偏微分方程,并采用数值计算方法获得该偏微分方程的迭代求解算法。同时在全散度变分CV模型中,增大拟合偏差项的权重系数,加大拟合偏差项在变分模型中的重要性。实验结果表明,全散度变分CV模型具有初始化敏感低、抗噪性强、鲁棒性高等优点。

关键词: 图像分割,CV模型,水平集,全散度

Abstract: The classical CV model is not completely suitable to segment the gray image which is intensity inhomogeneity,and has been disturbed by Gaussian noises with some variance.The variational CV model based on the total Bregman divergence was proposed and its iterative segmentation algorithm was presented.Firstly,the problems and disadvantages of the variational CV model segmentation method constructed by the Euclidean metric are analyzed.Secondly,compared with Euclidean metric,a figure shows the advantages of the total Bregman divergence that there is no connection with coordinate system in the distance calculation.Then,to reach the purpose of reducing noise sensitivity and enhance robustness of image segmentation,the data deviation term in CV model is built by the total Bregman divergence.Finally,Euler-Lagrange equation of this proposed variational model is obtained by variational method,and the variational model algorithm of the image segmentation is presented by numerical computation method.In addition,to accelerate the convergence rate,the weighting parameters of fitting terms should appropriately chose bigger value,and the importance of fitting items increases in variational model.The experimental results show that the proposed method is low sensitive to initialize contour curve,and has good anti-noise and robust performance.

Key words: Image segmentation,CV model,Level set method,Total Bregman divergence

王继策,吴成茂. 基于全散度的变分CV模型及其分割算法[J]. 计算机科学, 2015, 42(4): 306-310. https://doi.org/10.11896/j.issn.1002-137X.2015.04.063

WANG Ji-ce and WU Cheng-mao. Variational CV Model Based on Total Bregman Divergence and its Segmentation Algorithm[J]. Computer Science, 2015, 42(4): 306-310. https://doi.org/10.11896/j.issn.1002-137X.2015.04.063

参考文献

[1] Kass M,Witkin A,Terzopoulos D.Snakes:Active contour mo-dels[J].International Journal of Computer vision,1988,1(4):321-331
[2] Caselles V,Kimmel R,Sapiro G.Geodesic active contours[J].International Journal of Computer vision,1997,22(1):61-79
[3] Chan T F,Vese L A.Active contours without edges[J].IEEE Transactions on Image Processing,2001,10(2):266-277
[4] Osher S,Sethian J A.Fronts propagating with curvature de-pendent speed:algorithms based on Hamilton-Jacobi formulations[J].Journal of computational physics,1988,79(1):12-49
[5] Mumford D,Shah J.Optimal approximations by piecewisesmooth functions and associated variational problems[J].Communications on pure and applied mathematics,1989,42(5):577-685
[6] Tsai A,Yezzi Jr A,Willsky A S.Curve evolution implementa-tion of the Mumford-Shah functional for image segmentation,denoising,interpolation,and magnification[J].IEEE Transactions on Image Processing,2001,10(8):1169-1186
[7] Bertelli L,Sumengen B,Manjunath B S,et al.A variationalframework for multiregion pairwise- similarity- based image segmentation[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2008,30(8):1400-1414
[8] Krinidis S,Chatzis V.Fuzzy energy-based active contours[J].IEEE Transactions on Image Processing,2009,18(12):2747-2755
[9] Ben Salah M,Mitiche A,Ben Ayed I.Effective level set image segmentation with a kernel induced data term[J].IEEE Tran-sactions on Image Processing,2010,19(1):220-232
[10] Li C,Kao C Y,Gore J C,et al.Minimization of region-scalable fitting energy for image segmentation[J].IEEE Transactions on Image Processing,2008,17(10):1940-1949
[11] Wang L F,Yu Z Y,Pan C H.A unified level set framework utilizing parameter priors for medical image segmentation[J].Science China Information Sciences,2013,56(11):1-14
[12] Liu M.Total Bregman divergence,a robust diver-gence mea-sure,and its applications[D].University of Florida,2011
[13] Liu M,Vemuri B C,Amari S I,et al.Shape retrieval using hierarchical total Bregman soft clustering[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2012,34(12):2407-2419
[14] Li C,Huang R,Ding Z,et al.A level set method for image segmentation in the presence of intensity inhomogeneities with application to MRI[J].IEEE Transactions on Image Processing,2011,20(7):2007-2016
[15] Bregman L M.The relaxation method of finding the commonpoint of convex sets and its application to the solution of problems in convex programming[J].USSR Computational Mathematics and Mathematical Physics,1967,7(3):200-217
[16] Banerjee A,Merugu S,Dhillon I S,et al.Clustering with Bregman divergences[J].The Journal of Machine Learning Research,2005,6:1705-1749
[17] Liu M,Vemuri B C,Amari S I,et al.Total Bregman divergence and its applications to shape retrieval[C]∥2010 IEEE Confe-rence on Computer Vision and Pattern Recognition (CVPR).IEEE,2010:3463-3468
[18] Li C,Xu C,Gui C,et al.Distance regularized level set evolution and its application to image segmentation[J].IEEE Transactions on Image Processing,2010,19(12):3243-3254
[19] Li C,Xu C,Gui C,et al.Level set evolution without re-initialization:a new variational formulation[C]∥IEEE Computer Society Conference on Computer Vision and Pattern Recognition,2005(CVPR).IEEE,2005,1:430-436

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed