摘要
为减少阶梯效应,同时更好地利用图像本身的信息,提出一种结合分数阶全变差(FOTV)和非局部全变差(NLTV)模型的非盲去模糊图像重建方法。分别用FOTV和NLTV约束由全局梯度提取法(GGES)分解而成的平滑区和纹理区,建立图像非盲去模糊的正则化模型,分别采用交替方向乘子法(ADMM)和分裂Bregman操作符(BOS)算法求解两个子问题。充分的实验结果表明,该模型减少了平滑区的阶梯效应,更好地恢复了图像的纹理细节,验证了该模型的可行性和算法的有效性。
To reduce the staircase artifacts and use more information of the blurred image itself,an image reconstruction method of non-blind deblurring by combining the fractional order total variation(FOTV)and the non-local total variation(NLTV)models was proposed.The FOTV and NLTV regularization were used to constrain the flat and texture regions decomposed through global gradient extraction scheme(GGES),respectively,and a regularization optimization model of non-blind deblurring was constructed.Alternating direction method of multipliers(ADMM)was used to solve the FOTV sub-problems,while the algorithm of Bregman operator splitting(BOS)was used to solve NLTV sub-problems.Lots of experimental results and data demonstrate that the proposed method contains the advantages of FOTV and NLTV,it can not only reduce the staircase artifacts in flat regions,but also reconstruct more texture details of the blurred images.The feasibility and efficiency of the proposed method and corresponding algorithm are verified.
引文
[1]Goldstein T,Osher S.The split bregman method for L1-regularized problems[J].Siam Journal on Image Sciences,2009,2(2):323-343.
[2]Tao Min.Fast alternating direction method of multipliers for total-variation-based image restoration[J].Journal of Southeast University,2011,27(4):379-383.
[3]Zhang Xiaoqun,Burger M,Osher S.Bregman nonlocal regularization for deconvolution and sparse reconstruction[J].Siam Journal on Imaging Sciences,2010,3(3):253-276.
[4]Zhang Jun,Wei Zhihui.A class of fractional-order multi-scale variational models and alternating projection algorithm for image denoising[J].Applied Mathematical Modelling,2011,35(5):2516-2528.
[5]Ren Zemin,He Chuanjiang,Zhang Qifeng.Fractional order total variation regularization for image super-resolution[J].Signal Processing,2013,93(9):2408-2421.
[6]Xu Li,Lu Cewu,Xu Yi,et al.Image smoothing via L0gradient minimization[J].ACM Transactions on Graphics,2011,30(6):1-12.
[7]Tang Shu,Gong Weiguo,Li Weihong,et al.Non-blind image deblurring method by local and nonlocal total variation models[J].Signal Processing,2014,94(1):339-349.
[8]Zhang Zhengrong,Zhang Jun,Wei Zhihui,et al.Cartoontexture composite regularization based non-blind deblurring method for partly-textured blurred images with Poisson noise[J].Signal Processing,2015,116(C):127-140.
[9]HUANG Guo,XU Li,CHEN Qingli,et al.Research on image denosing based on space fractional partial differential equations[J].Journal of Sichuan University(Engineering Science Edition),2012,44(2):91-98(in Chinese)[黄果,许黎,陈庆利,等.基于空间分数阶偏微分方程的图像去噪模型研究[J].四川大学学报(工程科学版),2012,44(2):91-98.]
[10]Bioucas-Dias JM,Figueiredo MAT,Oliveira JP.Total variation-based image deconvolution:A majorization-minimization approach[C]//IEEE International Conference on Acoustics,2006.
[11]Karasulu B,Korukoglu S.Moving object detection and tracking by using annealed background subtraction method in videos:Performance optimization[J].Expert Systems with Applications,2012,39(1):33-43.
[12]Yang Junfeng,Zhang Yin,Yin Wotao.Fast alternating direction method for TVL1-L2signal reconstruction from partial Fourier data[J].Journal of Selected Topics in Signal Processing,2010,4(2):288-296.