博弈方法下的图像去噪与边界提取

doi:10.19665/j.issn1001-2400.2021.04.017

摘要/Abstract

摘要：

由于半二次正则化模型得到的边界过于模糊,去噪效果也不够理想,因此使用博弈的方法对半二次正则化模型加以改进,对图像同时进行去噪和边界提取。定义了两个参与者,采用经典的半二次正则化方法作为去噪的目标函数,选取比较新颖的全局稀疏梯度模型作为边界提取的目标函数。图像去噪与边界提取这两个参与者在一个博弈过程中交替迭代,将其收敛点作为纳什均衡点。实验结果说明,所提算法能够有效地改进半二次正则化模型,从而得到更好的去噪与边界提取效果。将提出的模型用于各种类别的图像,无论从数值结果还是视觉效果上,所提算法都得到了较好的结果。

关键词: 图像去噪, 边界提取, 半二次正则化, 全局稀疏梯度, 纳什均衡点

Abstract:

Half-quadratic regularization is a classical image denoising method.In removing image noise,the image boundary can be obtained.Since the boundary obtained by the half-quadratic regularization model is too fuzzy and the denoising effect is not ideal,the half-quadratic regularization model is improved by the game method,the image is denoised and the boundary is extracted simultaneously.Two participants are defined,with the classical half-quadratic regularization method used as the target function of denoising,and a relatively novel global sparse gradient model selected as the target function of boundary extraction.The two participants,image denoising and boundary extraction,iterate alternately in a game process,with their convergence points as the Nash equilibrium points.The proposed model is applied to various types of images,and the algorithm proposed can lead to good results in both numerical results and visual effects.Experimental results show that the proposed algorithm can effectively improve the half-quadratic regularization model,thus obtaining better denoising and boundary extraction effects.

Key words: image denoising, boundary extraction, half-quadratic regularization, global sparse gradient, nash equilibrium point

中图分类号:

TP391

乔鱼,冯象初. 博弈方法下的图像去噪与边界提取[J]. 西安电子科技大学学报, 2021, 48(4): 128-135.

QIAO Yu,FENG Xiangchu. Image denoising and boundary extraction based on game theory[J]. Journal of Xidian University, 2021, 48(4): 128-135.

图/表 8

图1

图2

表1

表2

图3

图4

图5

表3

参考文献 31

[1]	COHEN M C, HARSHA P. Designing Price Incentives in a Network with Social Interactions[J]. Manufacturing and Service Operations Management, 2020, 22(2):292-309. doi: 10.1287/msom.2018.0738
[2]	YANG Y, LIU L. Strategic Uncertainty,Coordination Failure and Emergence:A Game Theory Study on Agency Structure Interactions[J]. Journal for the Theory of SocialBehaviour, 2019, 49:402-420.
[3]	SHAO J, ZHAO Y, LIU X. Optimization of Test Resources for 3D SoC Based on Game Theory[J]. The Journal of China Universities of Posts and Telecommunications, 2019, 26(1):65-74.
[4]	SUMESH S, KRISHNA A, SUBRAMANIAN C. Game Theory Based Reasoning of Opposing Non functional Requirements using Interactor Dependencies[J]. Computer Journal, 2019, 62(11):1557-1583. doi: 10.1093/comjnl/bxy143
[5]	DANIELYAN A, KATKOVNIK V, EGIAZARIAN K. BM3D Frames and Variational Image Deblurring[J]. IEEE Transactions on Image Processing, 2012, 21(4):1715-1728. doi: 10.1109/TIP.2011.2176954
[6]	CHEN Y, WANG B, LIN W S, et al. Cooperative Peer-to-Peer Streaming:An Evolutionary Game Theoretic Approach[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2010, 20(10):1346-1357. doi: 10.1109/TCSVT.2010.2077490
[7]	KALLEL M, ABOULAICH R, HABBAL A, et al. A Nash Game Approach to Joint Image Restoration and Segmentation[J]. Applied Mathematical Modelling, 2014, 38(11-12):3038-3053. doi: 10.1016/j.apm.2013.11.034
[8]	JIN C, LI Q, JIN S. An Adaptive VPDE Image Denoising Model Based on Structure Tensor[J]. Multimedia Tools and Applications, 2019, 78(19):28331-28354. doi: 10.1007/s11042-019-07912-7
[9]	SERGIO A, SONIA B, JUAN R, et al. On Some New Variational Problems for Image Denoising[J]. Mathematical Methods in the Applied Sciences, 2019, 42(17):5881-5897. doi: 10.1002/mma.v42.17
[10]	PANG Z, ZHANG H, SHI B. Image Denoising Model Based on l p Directional Total Variation[J]. Journal of Beijing University of Aeronautics and Astronautics, 2019, 45(3):464.
[11]	MOHAMMED N. A Wavelet-based Method for MRI Liver Image Denoising[J]. Biomedizinische Technik/Biomedical Engineering, 2019, 64(6):699-709. doi: 10.1515/bmt-2018-0033
[12]	LI S Y, CHEN Y W, JIANG R X, et al. Image Denoising Via Multi-Scale Gated Fusion Network[J]. IEEE Access, 2019, 7:49392-49402. doi: 10.1109/Access.6287639
[13]	ZHANG K, ZUO W, ZHANG L. FFDNet:Toward A Fast and Flexible Solution for CNN Based Image Denoising[J]. IEEE Transactions on Image Processing, 2018, 27(9):4608-4622. doi: 10.1109/TIP.2018.2839891
[14]	ZHONG Y, LIU L, ZHAO D, et al. A Generative Adversarial Network for Image Denoising[J]. Multimedia Tools and Applications, 2019(2):1-13.
[15]	PERONA P, MALIK J. Scale-Space and Edge Detection Using Anisotropic Diffusion[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1990, 12(7):629-639. doi: 10.1109/34.56205
[16]	RUDIN L, OSHER S, FATEMI E. Nonlinear Total Variation Based Noise Removal Algorithms[J]. Physica D, 1992, 60(1-4):259-268. doi: 10.1016/0167-2789(92)90242-F
[17]	BUADES A, COLL B, MOREL J M. A Non-local Algorithm for Image Denoising[C]//Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2005, 2:60-65.
[18]	张瑞, 冯象初, 王斯琪, 等. 基于稀疏梯度场的非局部图像去噪算法[J]. 自动化学报, 2015, 41(9):1542-1552.
	ZHANG Rui, FENG Xiangchu, WANG Siqi, et al. A Sparse Gradients Field Based Image Denoising Algorithm Via Non-local Means[J]. Acta Automatica Sinica, 2015, 41(9):1542-1552.
[19]	GEMAN D, REYNOLDS G. Constrained Restoration and the Recovery of Discontinuities[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1992, 14(3):367-383. doi: 10.1109/34.120331
[20]	GREEN P J. Bayesian Reconstructions from Emission Tomography Data Using A Modified EM Algorithm[J]. IEEE Transactions on Medical Imaging, 1990, 9(1):84-93. doi: 10.1109/42.52985
[21]	CHARBONNIER P, BLANC F L, AUBERT G, et al. Deterministic Edge-preserving Regularization in Computed Imaging[J]. IEEE Transactions on Image Process, 1997, 6(2):298-311. doi: 10.1109/TIP.83
[22]	COHEN L D. Auxiliary Variables and Two-step Iterative Algorithms in Computer Vision Problems[J]. Journal of Mathematical Imaging and Vision, 1996, 6(1):59-83. doi: 10.1007/BF00127375
[23]	HEBERT T, LEAHY R. A Generalized EM Algorithm for 3-D Bayesian Reconstruction from Poisson Data Using Gibbs Priors[J]. IEEE Transactions on Medical Imaging, 1989, 8(2):194-202. doi: 10.1109/42.24868
[24]	LEITERITZ R. Capital Account Liberalization in Latin America:The Half-Life of a Powerful Economic Idea[J]. Optics and Spectroscopy, 2005, 98(3):336-340. doi: 10.1134/1.1890509
[25]	LI L H, PAN J S, LAI W S, et al. Learning A Discriminative Prior for Blind Image Deblurring[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE, 2018:6616-6625.
[26]	ZHAO Y B, ZHANG D M, WU Y F, et al. Supervised Descent Method for Weld Pool Boundary Extraction During Fiber Laser Welding Process[J]. China Welding, 2019, 1:6-10.
[27]	ZHANG Z Y, CAO Y F, ZHONG P Y, et al. An Edge-Boxes-Based Intruder Detection Algorithm for UAV Sense and Avoid System[J]. Transactions of Nanjing University of Aeronautics and Astronautics. 2019, 36(2):253-263.
[28]	李楠. SAR图像多目标边缘检测性能对比研究[J]. 舰船电子工程, 2019, 39(8):124-147.
	LI Nan. Comparative Study on the Performance of Multi-target Edge Detection in SAR Images[J]. Ship Electronic Engineering, 2019, 39(8):124-147.
[29]	ZHANG R, FENG X C, YANG L X, et al. Global Sparse Gradient Guided Variational Retinex Model for Image Enhancement[J]. Signal Processing:Image Communication, 2017, 58:270-281. doi: 10.1016/j.image.2017.08.008
[30]	张瑞, 冯象初, 杨丽霞, 等. 全局稀疏梯度耦合张量扩散的图像去噪模型[J]. 西安电子科技大学学报, 2017, 44(6):150-155.
	ZHANG Rui, FENG Xiangchu, YANG Lixia, et al. Image Denoising Model Based on Global Sparse Gradient Coupled Tensor Diffusion[J]. Journal of Xidian University, 2017, 44(6):150-155.
[31]	COMBETTES P L, WAJS V R. Signal Recovery by Proximal Forward-Backward Splitting[J]. Multiscale Modeling and Simulation, 2005, 4(4):1168-1200. doi: 10.1137/050626090

算法	house	mon	brain	lena	couple	man	hill	camera
GRHQ	31.04	29.56	35.15	30.82	27.62	28.83	28.79	27.10
GMHQ	27.08	26.10	28.63	26.02	24.46	25.30	25.04	26.38
HLHQ	30.41	29.76	32.75	29.71	27.61	28.33	28.20	28.28
HSHQ	31.05	29.56	35.19	30.86	27.67	28.88	28.84	27.12
PM	27.93	27.89	34.26	28.43	25.57	26.77	27.25	24.78
NLM	28.56	28.06	29.60	28.43	27.26	27.52	27.56	27.20
SGNLM	31.49	30.91	33.17	30.26	26.92	27.80	27.68	28.33
GSGTD	31.08	28.36	35.40	31.08	28.79	29.47	29.49	27.54
HQGSG	31.55	30.25	35.41	31.45	28.58	29.58	29.50	28.31

算法	house	mon	brain	lena	couple	man	hill	camera
GRHQ	0.766	0.875	0.889	0.894	0.847	0.863	0.848	0.848
GMHQ	0.698	0.763	0.698	0.752	0.680	0.706	0.653	0.719
HLHQ	0.754	0.839	0.856	0.877	0.845	0.854	0.833	0.772
HSHQ	0.765	0.873	0.890	0.895	0.850	0.868	0.851	0.760
PM	0.692	0.823	0.871	0.855	0.798	0.823	0.815	0.684
NLM	0.550	0.609	0.795	0.845	0.873	0.866	0.863	0.570
SGNLM	0.840	0.906	0.866	0.822	0.706	0.731	0.681	0.814
GSGTD	0.813	0.798	0.879	0.818	0.775	0.784	0.760	0.724
HQGSG	0.771	0.873	0.898	0.902	0.878	0.889	0.876	0.780

图像	GRHQ	GMHQ	HLHQ	HSHQ	HQGSG
house	36.45	135.91	132.76	33.37	68.30
man	199.36	717.84	704.49	174.12	325.10