JIA Li-na1,2, JIAO Feng-yuan2, LIU Rui-qiang3, GUI Zhi-guo2
（1. Department of Electronic Information Engineering, Shanxi University, Taiyuan 030013, China； 2. Shanxi Provincial Key Laboratory for Biomedical Imaging and Big Data, North University of China, Taiyuan 030051, China；3. School of Electronics and Information Engineering, Harbin Institute of Technology, Harbin 150001, China）
Abstract： Classic non-local means (CNLM) algorithm uses the inherent self-similarity in images for noise removal. The denoised pixel value is estimated through the weighted average of all the pixels in its non-local neighborhood. In the CNLM algorithm, the differences between the pixel value and the distance of the pixel to the center are both taken into consideration to calculate the weighting coefficients. However, the Gaussian kernel cannot reflect the information of edge and structure due to its isotropy, and it has poor performance in flat regions. In this paper, an improved non-local means algorithm based on local edge direction is presented for image denoising. In edge and structure regions, the steering kernel regression (SKR) coefficients are used to calculate the weights, and in flat regions the average kernel is used. Experiments show that the proposed algorithm can effectively protect edge and structure while removing noises better when compared with the CNLM algorithm.
Key words： image denoising; neighborhood filter; non-local means (NLM); steering kernel regression (SKR)
CLD number： TN911.73 Document code： A
Article ID： 1674-8042（2019）03-0236-05 doi： 103969/jissn1674-80422019-03-006
References
［1］Aubert G, Kornprobst P. Mathematical problems in image processing. New York: Springer Verlag, 2002.
［2］Gonzalez R C, Woods R E. Digital image processing. Englewood: Prentice Hall, 2007.
［3］Buades A, Coll B, Morel J M. Neighborhood filters and pde’s. Journal Numerische Mathematik, 2006, 105(1): 1-34.
［4］Buades A, Coll B, Morel J M. A non-local algorithm for image denoising. In: Proceedings of CVPR’05, San Diego, IEEE Press, 2005: 60-65.
［5］Buades A, Coll B, Morel J M. A review of image denoising algorithms, with a new one. Multiscale Modeling and Simulation, 2005, 4(2): 490-530.
［6］Katkovnik V, Foi A, Egiazarian K. From local kernel to nonlocal multiple-model image denoising. International Journal of Computer Vision, 2010, 86(1): 1-32.
［7］Zhang H, Yang J, Zhang Y, et al., Image and video restorations via nonlocal kernel regression. IEEE Transactions on Cybernetics, 2013, 43(3): 1035-1046.
［8］Liu Y L, Wang J, Chen X, et al. A robust and fast non-local means algorithm for image denoising. Journal of Computer Science and Technology. 2008, 23(2): 270-279.
［9］Chen Y, Yang Z, Hu Y, et al. Thoracic low-dose CT image processing using an artifact suppressed large-scale nonlocal means. Physics in Medicine & Biology, 2012, 57(9): 2667-2688.
［10］Zhang Q, Gui Z, Liu Y, et al. Adaptive non-local means denoising algorithm for medical image. Computer Engineering, 2012, 38(7): 182-184.
［11］Li Z, Yu L, Trzasko J D, et al. Adaptive nonlocal means filtering based on local noise level for CT denoising. Medical Physics, 2014, 41(1): 011908.
［12］Zhang Y, Lu H, Rong J, et al. Adaptive non-local means on local principle neighborhood for noise/artifacts reduction in low-dose CT images. Medical Physics, 2017, 44(9): e230-e241.
［13］Jia L, Zhang Q, Liu Y, et al. A two-step denoising method for low dose computed tomography image via morphological component analysis and non-local means. Journal of Medical Imaging and Health Informatics, 2019, 9: 140-147.
［14］Wang X T, Shi G M, Niu Y, et al. Robust adaptive directional lifting wavelet transform for image denoising. IET Image Processing, 2011, 5(3): 249-260.
［15］Takeda H, Farsiu S, Milanfar P. Kernel regression for image processing and reconstruction. IEEE Transactions on Image Processing, 2007, 16(2): 349-366.
［16］Chatterjee P, Milanfar P. Clustering-based denoising with locally learned dictionaries. IEEE Transactions on Image Processing, 2009, 18(7): 1438-1451.
［17］ Chatterjee P, Milanfar P. Patch-based near-optimal image denoising. IEEE Transactions on Image Processing, 2012, 21(4): 1635-1649.
［18］Wang Z, Bovik A C, Sheikh H R, et al. Image quality assessment: From error visibility to structural similarity. IEEE Transactions on Image Processing, 2004, 13(4): 600-612.
基于局部边沿方向的非局部均值图像去噪算法
贾丽娜1,2, 焦枫媛2， 刘瑞强3， 桂志国2
（1. 山西大学 电子信息工程系， 山西 太原 030013； 2. 中北大学 山西省医学成像与影像大数据重点实验室， 山西 太原 030051；3. 哈尔滨工业大学 电子与信息工程学院， 黑龙江 哈尔滨 150001）
摘要:经典的非局部均值（Classic non-local means, CNLM）算法利用图像中自有的自相似性对噪声进行去除。 去噪像素值通过其非局部邻域中所有像素的加权平均而得到。 在CNLM算法中， 计算权重系数的过程中同时考虑了像素值和该像素点距离中心像素点的距离之间的差异。 然而， 由于各向同性， 高斯核不能反应边沿和结构信息， 并且高斯核在平坦区域表现欠佳。 本文提出了一种改进的基于局部边沿方向的非局部均值图像去噪算法。 在边沿和结构区域使用旋转核回归系数（Steering kernel regression, SKR）计算权重系数，在平坦区域使用平均核计算权重系数。 实验结果表明， 与CNLM算法相比， 提出的算法可以在有效地保护边沿和结构的同时更好地去除噪声。
关键词:图像去噪； 邻域滤波； 非局部均值（NLM）； 旋转核回归（SKR）
引用格式：JIA Li-na, JIAO Feng-yuan, LIU Rui-qiang, et al. Local edge direction based non-local means for image denoising. Journal of Measurement Science and Instrumentation, 2019, 10（3）： 236-240. ［doi： 103969／jissn1674-8042201903006］
[full text view]
