基于局部期望最大化注意力的图像降噪

李泽田; 雷志春

doi:10.3788/YJYXS20203504.0350

您当前的位置：

首页 >

文章列表页 >

基于局部期望最大化注意力的图像降噪

图像处理 | 更新时间：2020-07-05

- 基于局部期望最大化注意力的图像降噪
- Local expectation-maximization attention network for image denoising
- 液晶与显示 2020年35卷第4期页码：350-359
- 作者机构：
  
  天津大学微电子学院, 天津 300072
- 作者简介：
  
  [ "李泽田(1991-), 男, 河北邢台人, 硕士, 2015年于河北工业大学大学获得硕士学位, 主要从事图像降噪, 计算机视觉的研究。E-mail:lizetian527@tju.edu.cn" ]
  [ "雷志春(1964-), 男, 河北迁安人, 博士, 教授, 1998年于德国多特蒙德大学获得博士学位, 主要从事高动态范围与宽色域算法研究。E-mail:zclei@tju.edu.cn" ]
- 基金信息：
  
  天津市科技支撑重点项目(16YFZCGX00760)
- DOI：10.3788/YJYXS20203504.0350
  中图分类号： TP391.1
- 收稿日期：2019-09-13，
  
  修回日期：2019-11-04，
  
  录用日期：2019-11-4，
  
  纸质出版日期：2020-04-05
- 稿件说明：
移动端阅览
李泽田, 雷志春. 基于局部期望最大化注意力的图像降噪[J]. 液晶与显示, 2020,35(4):350-359.

Ze-tian LI, Zhi-chun LEI. Local expectation-maximization attention network for image denoising[J]. Chinese journal of liquid crystals and displays, 2020, 35(4): 350-359.
李泽田, 雷志春. 基于局部期望最大化注意力的图像降噪[J]. 液晶与显示, 2020,35(4):350-359. DOI： 10.3788/YJYXS20203504.0350.

Ze-tian LI, Zhi-chun LEI. Local expectation-maximization attention network for image denoising[J]. Chinese journal of liquid crystals and displays, 2020, 35(4): 350-359. DOI： 10.3788/YJYXS20203504.0350.

摘要

图像降噪算法生成的图像往往会因丢失高频信息而造成细节模糊，并且随着网络的加深，噪声在特征域会不断累加。为了恢复图像降噪中生成图像的高频细节，增强网络的拟合能力和泛化能力，减少噪声在特征域的传播，本文提出了一种基于局部期望最大化注意力机制的降噪网络。网络以经典的自动编码器为基本模块以达到足够的深度，实现足够大的感受野，并滤除不必要的低频信息，促进网络学习高频信息。局部期望最大化注意力模块利用局部期望最大化算法迭代出紧凑的局部基，并在该局部基上应用注意力机制，大大降低网络复杂度。大量实验表明，本文提出的局部期望最大化注意力网络可以恢复更多的高频细节，在基准测试集BSD68上，PSNR值比当前最先进算法N

Net提升约0.13 dB，并有更好的视觉质量。

Abstract

Images generated by image denoising algorithms often result in missing details due to loss of high frequency information. In order to restore the high-frequency details of the reconstructed images and enhance the representation ability of the network

a noise removing network based on local expectation-maximization attention(LEMA) is proposed. The network uses the classic autoencoder as the basic module to achieve sufficient depth

so as to filter out unnecessary low-frequency information and promote network learning high-frequency information. The proposed channel and spatial attention module adaptively learns informativel features by computing the interdependence of image channels and pixel spatial positions. A series of experiments show that the LEMA can recover more high frequency details. On the benchmark BSD68

the PSNR value is about 0.13 dB higher than the state-of-the-art algorithm N

Net

and has better visual quality.

关键词

Keywords

references

F HEIDE , M STEINBERGER , Y T TSAI , 等 . FlexISP:a flexible camera image processing framework . ACM Transactions on Graphics , 2014 . 33 ( 6 ): 231 .

ZHANG K, ZUO W M, GU S H, et al . Learning deep CNN denoiser prior for image restoration[C] // Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition . Honolulu, HI, USA: IEEE, 2017: 2808-2817.

肖启阳 , 李健 , 吴思进 , 等 . 基于能量的DSPI相位图正余弦降噪 . 光学精密工程 , 2018 . 26 ( 4 ): 936 - 943 . http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201804023 http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201804023 .

Q Y XIAO , J LI , S J WU , 等 . Denoising of DSPI phase map using sine-cosine filtering and signal energy . Optics and Precision Engineering , 2018 . 26 ( 4 ): 936 - 943 . http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201804023 http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201804023 .

陈真诚 , 吴贤亮 , 赵飞骏 . EEMD结合小波阈值的光电容积脉搏波信号降噪方法及实现 . 光学精密工程 , 2019 . 27 ( 6 ): 1327 - 1334 .

Z C CHEN , X L WU , F J ZHAO . Denoising and implementation of photoplethysmography signal based on EEMD and wavelet threshold . Optics and Precision Engineering , 2019 . 27 ( 6 ): 1327 - 1334 .

孙京阳 , 喻春雨 , 董仕佳 . 非局部均值噪声预测的独立成分分析降噪研究 . 光学精密工程 , 2018 . 26 ( 2 ): 511 - 516 . http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=gxjmgc201802030 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=gxjmgc201802030 .

J Y SUN , C Y YU , S J DONG . Noise reduction of independent component analysis based on NLmeans noise prediction . Optics and Precision Engineering , 2018 . 26 ( 2 ): 511 - 516 . http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=gxjmgc201802030 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=gxjmgc201802030 .

丁明君 , 习江涛 , 李光旭 , 等 . 基于三波长相移轮廓术的表面测量降噪 . 光学精密工程 , 2017 . 25 ( 10s ): 297 - 303 .

M J DING , J T XI , G X LI , 等 . Denoising method for shape measurement based on three wavelength phase shift profilometry . Optics and Precision Engineering , 2017 . 25 ( 10s ): 297 - 303 .

BUADES A, COLL B, MOREL J M. A non-local algorithm for image denoising[C]// Proceedings of 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition . San Diego, USA: IEEE, 2005: 60-65.

S ROTH , M J BLACK . Fields of experts . International Journal of Computer Vision , 2009 . 82 ( 2 ): 205 - 229 . DOI: 10.1007/s11263-008-0197-6 http://doi.org/10.1007/s11263-008-0197-6 .

S CHO , S LEE . Fast motion deblurring . ACM Transactions on Graphics , 2009 . 28 ( 5 ): 145 http://d.old.wanfangdata.com.cn/Periodical/hwyjggc201704038 http://d.old.wanfangdata.com.cn/Periodical/hwyjggc201704038 .

XU L, JIA J Y. Two-phase kernel estimation for robust motion deblurring[C] // Proceedings of the 11 th European Conference on Computer Vision . Heraklion, Crete, Greece: Springer, 2010: 157-170.

唐玲 , 陈明举 . 基于图像局部方向特性的自适应全变分去噪模型 . 液晶与显示 , 2016 . 31 ( 5 ): 477 - 483 . http://yjyxs.com/CN/abstract/abstract10017.shtml http://yjyxs.com/CN/abstract/abstract10017.shtml .

L TANG , M J CHEN . Adaptive directional total variation denoising model based on image local direction . Chinese Journal of Liquid Crystals and Displays , 2016 . 31 ( 5 ): 477 - 483 . http://yjyxs.com/CN/abstract/abstract10017.shtml http://yjyxs.com/CN/abstract/abstract10017.shtml .

GU S H, ZHANG L, ZUO W M, et al . Weighted nuclear norm minimization with application to image denoising[C]// Proceedings of 2014 IEEE Conference on Computer Vision and Pattern Recognition . Columbus, USA: IEEE, 2014: 2862-2869.

K ZHANG , W M ZUO , Y J CHEN , 等 . Beyond a gaussian denoiser:residual learning of deep CNN for image denoising . IEEE Transactions on Image Processing , 2017 . 26 ( 7 ): 3142 - 3155 . DOI: 10.1109/TIP.2017.2662206 http://doi.org/10.1109/TIP.2017.2662206 .

LIU P J, ZHANG H Z, ZHANG K, et al . Multi-level wavelet-CNN for image restoration[C]// Proceedings of the IEEE / CVF Conference on Computer Vision and Pattern Recognition Workshops . Salt Lake City, USA: IEEE, 2018: 88601-88609.

K ZHANG , W M ZUO , L ZHANG . FFDNet:toward a fast and flexible solution for CNN-based image denoising . IEEE Transactions on Image Processing , 2018 . 27 ( 9 ): 4608 - 4622 . DOI: 10.1109/TIP.2018.2839891 http://doi.org/10.1109/TIP.2018.2839891 .

PLÖTZ T, ROTH S. Neural nearest neighbors networks[C]// Proceedings of the 32nd Conference on Neural Information Processing Systems . Montréal, Canada: NeruIPS, 2018: 1087-1098.

LIAO F Z, LIANG M, DONG Y P, et al . Defense against adversarial attacks using high-level representation guided denoiser[C]// Proceedings of 2018 IEEE / CVF Conference on Computer Vision and Pattern Recognition . Salt Lake City, USA: IEEE, 2018: 1778-1787.

VASWANI A, SHAZEER N, PARMAR N, et al . Attention is all you need[C]// Proceedings of 31st Conference on Neural Information Processing Systems . Long Beach, CA, USA: NeurIPS, 2017: 5999-6009.

WANG X L, GIRSHICK R, GUPTA A, et al . Non-local neural networks[C]// Proceedings of 2018 IEEE / CVF Conference on Computer Vision and Pattern Recognition . Salt Lake City, USA: IEEE, 2018: 7794-7803.

ZHANG Y L, LI K P, LI K, et al . Image super-resolution using very deep residual channel attention networks[C]// Proceedings of European Conference on Computer Vision .Munich, Germany: Springer, 2018: 294-310.

LI X, ZHONG Z S, WU J L, et al . Expectation-maximization attention networks for semantic segmentation[C]// Proceedings of 2019 International Conference on Computer Vision . Seoul, Korea: ICCV, 2019.

MAO X J, SHEN C H, YANG Y B. Image restoration using very deep convolutional encoder-decoder networks with symmetric skip connections[C]// Proceedings of the 30 th International Conference on Neural Information Processing Systems . Barcelona, Spain: Curran Associates Inc, 2016: 2802-2810.

BURGER H C, SCHULER C J, HARMELING S. Image denoising: can plain neural networks compete with BM3D?[C]// Proceedings of 2012 IEEE Conference on Computer Vision and Pattern Recognition .Providence, RI, USA: IEEE, 2012: 2392-2399.

XIE J Y, XU L L, CHEN E H. Image denoising and inpainting with deep neural networks[C] // Proceedings of the 25 th Neural Information Processing Systems . Lake Tahoe, Nevada: Curran Associates Inc, 2012: 341-349.

CHO K. Simple sparsification improves sparse denoising autoencoders in denoising highly corrupted images[C]// Proceedings of the 30 th International Conference on Machine Learning . Atlanta, USA: PMLR, 2013: 432-440.

K DABOV , A FOI , V KATKOVNIK , 等 . Image denoising by sparse 3-D transform-domain collaborative filtering . IEEE Transactions on Image Processing , 2007 . 16 ( 8 ): 2080 - 2095 . DOI: 10.1109/TIP.2007.901238 http://doi.org/10.1109/TIP.2007.901238 .

M ELAD , M AHARON . Image denoising via sparse and redundant representations over learned dictionaries . IEEE Transactions on Image Processing , 2006 . 15 ( 12 ): 3736 - 3745 . DOI: 10.1109/TIP.2006.881969 http://doi.org/10.1109/TIP.2006.881969 .

P DEMPSTERA , N M LAIRD , D B RUBIN . Maximum likelihood from incomplete data via the EM algorithm . Journal of the Royal Statistical Society:Series B (Methodological) , 1977 . 39 ( 1 ): 1 - 22 . DOI: 10.1111/j.2517-6161.1977.tb01600.x http://doi.org/10.1111/j.2517-6161.1977.tb01600.x .

于长淞 , 方超 . 基于小波变换的ESPI图像去噪及边缘提取 . 液晶与显示 , 2011 . 26 ( 6 ): 818 - 822 . http://yjyxs.com/CN/abstract/abstract9007.shtml http://yjyxs.com/CN/abstract/abstract9007.shtml .

C S YU , C FANG . ESPI image denoising and edge extraction based on wavelet transform . Chinese Journal of Liquid Crystals and Displays , 2011 . 26 ( 6 ): 818 - 822 . http://yjyxs.com/CN/abstract/abstract9007.shtml http://yjyxs.com/CN/abstract/abstract9007.shtml .

WOO S, PARK J, LEE J Y, et al . CBAM: convolutional block attention module[C]// Proceedings of the 15 th European Conference on Computer Vision . Munich, Germany: Springer, 2018: 3-19.

JOHNSON J, ALAHI A, LI F F. Perceptual losses for real-time style transfer and super-resolution[C]// Proceedings of the 14 th European Conference on Computer Vision .Amsterdam, The Netherlands: Springer, 2016: 694-711. http://link.springer.com/chapter/10.1007/978-3-319-46475-6_43

SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[J]. arXiv preprint arXiv: 1409.1556, 2014.

TIMOFTE R, AGUSTSSON E, VAN GOOL L, et al . NTIRE 2017 challenge on single image super-resolution: methods and results[C]// Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops .Honolulu, USA: IEEE, 2017: 1110-1121.

MARTIN D, FOWLKES C, TAL D, et al . A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics[C]// Proceedings of the 8 th IEEE International Conference on Computer Vision . Vancouver, Canada: IEEE, 2001: 416-423.

HUANG J B, SINGH A, AHUJA N. Single image super-resolution from transformed self-exemplars[C]// Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition . Boston, USA: IEEE, 2015: 5197-5206. http://www.researchgate.net/publication/275523282_Single_Image_Super-resolution_from_Transformed_Self-Exemplars

Z WANG , A C BOVIK , H R SHEIKH , 等 . Image quality assessment:from error visibility to structural similarity . IEEE Transactions on Image Processing , 2004 . 13 ( 4 ): 600 - 612 . DOI: 10.1109/TIP.2003.819861 http://doi.org/10.1109/TIP.2003.819861 .

KINGMA D P, BA J. Adam: a method for stochastic optimization[C]// Proceedings of the 3 rd International Conference for Learning Representations . San Diego, USA: ICLR, 2015.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

暂无数据