Please wait a minute...
 
Remote Sensing for Land & Resources    2018, Vol. 30 Issue (2) : 114-124     DOI: 10.6046/gtzyyg.2018.02.16
|
Research on image denoising algorithm of joint bilateral filter and wavelet threshold shrinkage
Shangwang LIU1,2(), Liuyang GAO1,2, Bo WANG1,2
1.College of Computer and Information Engineering, Henan Normal University, Xinxiang 453007, China
2.Engineering Lab of Intelligence Business and Internet of Things, Xinxiang 453007, China
Download: PDF(10300 KB)   HTML
Export: BibTeX | EndNote | Reference Manager | ProCite | RefWorks    
Abstract   Aim

ing at overcoming the shortcomings of existing denoising algorithms, such as the poor denoising capability, the noise error evaluation, and the damaging of the image edge and texture details, this paper proposes an image denoising algorithm of joint bilateral filter and wavelet threshold shrinkage. Firstly, the original noise image is divided into high-contrast and low-contrast layers by bilateral filter. Secondly, different appropriate filters are employed for different hierarchical layers. i.e., the bilateral filter and wavelet threshold shrinkage are adopted for high-contrast and low-contrast layers, respectively. Finally, the final denoising image is obtained by integrating high-contrast with low-contrast layers’ denoising images, which suppresses noises and at the same time enhances the image more efficiently. Experimental results show that peak signal to noise ratio (PSNR) of this method reaches 40.99 dB, which is higher than the ratio of non-local means filter, bilateral filter, wavelet threshold shrinkage and partial differential equation algorithms by 7.79%, 3.56%, 11.22% and 1.91%, respectively. Moreover, the proposed algorithm can not only remove the noises efficiently but also preserve the image edge and texture details very well.
Keywords bilateral filter      wavelet threshold shrinkage      peak signal to noise ratio      image edge      texture     
:  TP79  
Issue Date: 30 May 2018
Service
E-mail this article
E-mail Alert
RSS
Articles by authors
Shangwang LIU
Liuyang GAO
Bo WANG
Cite this article:   
Shangwang LIU,Liuyang GAO,Bo WANG. Research on image denoising algorithm of joint bilateral filter and wavelet threshold shrinkage[J]. Remote Sensing for Land & Resources, 2018, 30(2): 114-124.
URL:  
https://www.gtzyyg.com/EN/10.6046/gtzyyg.2018.02.16     OR     https://www.gtzyyg.com/EN/Y2018/V30/I2/114
Fig.1  Flowchart of algorithm in this paper
Fig.2  BF denoising
Fig.3  Wavelet threshold denoising
Fig.4  Results of five denoising algorithms
Fig.5  Edge detection of five denoising results
Fig.6  Gray histograms of five denoising results
算法 图像编号 高斯噪声密度 平均值
0.005 0.010 0.015 0.200
NLM ① 30.03 29.72 29.24 28.71 29.43
② 28.35 28.07 27.78 27.32 27.88
③ 28.28 28.00 27.73 27.31 27.83
④ 28.93 28.67 28.29 27.81 28.43
⑤ 29.34 29.01 28.53 28.05 28.73
⑥ 27.53 27.28 26.96 26.63 27.05
BF ① 42.06 40.75 39.42 37.81 40.01
② 39.16 38.27 37.36 37.02 37.95
③ 33.11 32.00 31.47 28.93 31.34
④ 37.77 36.35 35.02 33.14 35.57
⑤ 36.27 35.81 33.32 32.23 34.41
⑥ 34.39 33.12 32.25 30.01 32.44
小波阈值收缩 ① 20.60 18.42 16.99 15.29 17.83
② 20.13 18.52 16.68 16.01 17.84
③ 24.07 23.15 22.46 21.82 22.88
④ 27.69 25.87 24.60 23.68 25.46
⑤ 28.57 26.47 25.10 23.67 25.67
⑥ 27.43 25.72 24.57 23.66 25.34
PDE ① 41.84 39.53 38.04 36.96 39.10
② 40.74 38.12 36.42 35.33 37.65
③ 40.35 37.35 35.32 33.78 36.70
④ 40.96 38.31 36.66 35.45 37.85
⑤ 40.95 38.44 36.92 35.87 38.05
⑥ 41.00 38.29 36.46 35.13 37.72
本文算法 ① 46.56 41.76 39.63 38.36 41.58
② 46.31 41.13 38.51 37.88 40.96
③ 45.95 40.86 38.69 37.16 40.67
④ 46.13 41.23 39.16 37.18 40.93
⑤ 46.01 41.18 39.03 37.56 40.95
⑥ 46.60 41.59 38.97 36.24 40.85
Tab.1  PSNR in different levels Gaussian noise(dB)
算法 图像编号 椒盐噪声密度 平均值
0.1 0.2 0.3 0.4
NLM ① 28.63 27.94 27.63 26.82 27.76
② 26.72 25.58 24.62 23.45 25.09
③ 26.53 25.43 24.54 23.21 24.93
④ 26.85 25.71 24.81 23.62 25.25
⑤ 25.48 24.68 23.43 22.89 24.12
⑥ 25.64 24.20 22.85 22.03 23.68
BF ① 40.17 38.72 37.53 36.80 38.31
② 37.28 36.67 35.73 34.21 35.97
③ 31.34 29.47 28.64 27.32 29.19
④ 35.73 33.63 32.71 33.43 33.88
⑤ 33.58 31.57 30.46 29.38 31.25
⑥ 29.97 27.51 26.37 25.26 27.28
小波阈值收缩 ① 18.79 16.54 14.71 13.23 15.82
② 18.34 16.63 14.68 13.82 15.87
③ 22.47 21.26 20.51 19.32 20.89
④ 25.51 24.27 23.65 22.07 23.88
⑤ 26.37 24.50 23.33 22.72 24.23
⑥ 25.18 23.62 22.41 21.55 23.19
PDE ① 39.79 37.62 36.37 35.28 37.27
② 38.67 36.37 35.40 34.39 36.21
③ 38.21 35.33 34.58 33.42 35.39
④ 38.72 36.51 35.49 34.41 36.28
⑤ 38.84 36.63 35.38 34.32 36.29
⑥ 39.20 36.47 35.35 34.20 36.31
本文算法 ① 45.42 42.65 40.07 38.83 41.74
② 45.17 42.21 38.97 37.60 40.99
③ 44.82 41.76 38.42 37.00 40.50
④ 45.03 41.59 40.01 38.21 41.21
⑤ 44.96 40.52 39.21 37.82 40.63
⑥ 45.00 40.63 38.87 36.97 40.37
Tab.2  PSNR in different densities impulse noise(dB)
Fig.7  Results of different denoising algorithms
指标 算法 高斯噪声密度
0.2 0.4 0.6 0.8 1.0
PSNR NLM 18.24 17.05 16.04 15.37 13.72
BF 22.47 21.36 20.31 19.44 17.51
小波阈值收缩 15.56 14.21 13.54 12.83 10.94
PDE 23.60 22.13 21.33 20.04 18.45
本文算法 26.17 25.84 24.25 23.05 21.57
EPI NLM 0.246 0.217 0.206 0.191 0.178
BF 0.394 0.374 0.362 0.351 0.337
小波阈值收缩 0.187 0.173 0.164 0.152 0.141
PDE 0.422 0.408 0.391 0.375 0.340
本文算法 0.541 0.524 0.510 0.499 0.476
Tab.3  PSNR and EPI in different levels Gaussian noise(dB)
指标 算法 椒盐噪声密度
0.1 0.3 0.5 0.7 0.9
PSNR NLM 16.37 15.28 14.73 12.38 11.57
BF 20.35 19.47 18.38 17.56 16.39
小波阈值收缩 13.64 12.72 11.53 10.20 9.48
PDE 21.46 20.52 19.37 18.04 17.21
本文算法 24.58 22.88 22.34 21.54 20.37
EPI NLM 0.234 0.205 0.194 0.187 0.172
BF 0.383 0.372 0.360 0.345 0.321
小波阈值收缩 0.172 0.158 0.146 0.135 0.120
PDE 0.413 0.392 0.381 0.365 0.334
本文算法 0.531 0.517 0.502 0.489 0.473
Tab.4  PSNR and EPI in different densities impulse noise (dB)
[1] Knaus C, Zwicker M. Dual-domain image denoising [C]//Proceedings of 20th IEEE International Conference on Image Processing.Melbourne:IEEE, 2013: 440-444.
[2] 张倩 . 基于双重离散小波变换的遥感图像去噪算法[J]. 国土资源遥感, 2015,27(4):14-20.doi: 10.6046/gtzyyg.2015.04.03.
doi: 10.6046/gtzyyg.2015.04.03
[2] Zhang Q . Remote sensing image de-noising algorithm based on double discrete wavelet transform[J]. Remote Sensing for Land and Resources, 2015,27(4):14-20.doi: 10.6046/gtzyyg.2015.04.03.
[3] Huang Q G, Hao B Y, Chang S . Adaptive digital ridgelet transform and its application in image denoising[J]. Digital Signal Processing, 2016,52:45-54.
doi: 10.1016/j.dsp.2016.02.004 url: http://linkinghub.elsevier.com/retrieve/pii/S1051200416000324
[4] 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:IEEE, 2005: 60-65.
[5] 谭茹, 李婷婷, 李伟伟 , 等. 图像去噪的自适应非局部均值滤波方法[J]. 小型微型计算机系统, 2014,35(1):137-141.
doi: 10.3969/j.issn.1000-1220.2014.01.028 url: http://d.wanfangdata.com.cn/Periodical/xxwxjsjxt201401028
[5] Tan R, Li T T, Li W W , et al. Adaptive non-local means filtering method for image denoising[J]. Journal of Chinese Computer Systems, 2014,35(1):137-141.
[6] 黄智, 付兴武, 刘万军 . 混合相似性权重的非局部均值去噪算法[J]. 计算机应用, 2016,36(2):556-562.
doi: 10.11772/j.issn.1001-9081.2016.02.0556
[6] Huang Z, Fu X W, Liu W J . Non-local means denoising algorithm with hybrid similarity weight[J]. Journal of Computer Applications, 2016,36(2):556-562.
[7] 周兵, 韩媛媛, 徐明亮 , 等. 快速非局部均值图像去噪算法[J]. 计算机辅助设计与图形学学报, 2016,28(8):1260-1268.
[7] Zhou B, Han Y Y, Xu M L , et al. A fast non-local means image denoising algorithm[J]. Journal of Computer-Aided Design and Computer Graphics, 2016,28(8):1260-1268.
[8] Tomasi C, Manduchi R. Bilateral filtering for gray and color images [C]//Proceedings of the 6th International Conference on Computer Vision. Bombay:IEEE, 2010: 839-846.
[9] 杨学志, 徐勇, 方静 , 等. 结合区域分割和双边滤波的图像去噪新算法[J]. 中国图象图形学报, 2012,17(1):40-48.
doi: 10.11834/jig.20120106 url: http://d.wanfangdata.com.cn/Periodical/zgtxtxxb-a201201006
[9] Yang X Z, Xu Y, Fang J , et al. New filter based on region segmentation and bilateral filtering[J]. Journal of Image and Graphics, 2012,17(1):40-48.
[10] Ramesh S.An efficient approach for removal of universal noise using adaptive based switching bilateral filter[C]//Proceedings of 2012 International Conference on Advances in Engineering,Science and Management.Nagapattinam:IEEE, 2012: 462-467.
[11] 袁华, 庞建铿, 莫建文 . 基于噪声分类的双边滤波点云去噪算法[J]. 计算机应用, 2015,35(8):2305-2310.
doi: 10.11772/j.issn.1001-9081.2015.08.2305 url: http://www.cqvip.com/QK/94832X/201508/665663330.html
[11] Yuan H, Pang J K, Mo J W . Denoising algorithm for bilateral filtered point cloud based on noise classification[J]. Journal of Computer Applications, 2015,35(8):2305-2310.
[12] 杨燕 . 基于变分偏微分方程的图像去噪及其快速算法[D]. 南京:南京邮电大学, 2015.
[12] Yang Y . Image Denoising Based on Calculus of Variations and Partial Differential Equations,and Its Fast Algorithm[D]. Nanjing: Nanjing University of Posts and Telecommunications, 2015.
[13] Halim S A, Ibrahim A, Sulong T N N T,et al.Fourth-order partial differential equation noise removal on welding images[J]. AIP Conference Proceedings, 2015,1682(1):020050.
doi: 10.1063/1.4932459 url: http://aip.scitation.org/doi/abs/10.1063/1.4932459
[14] 芦碧波, 李阳, 王永茂 . 结合松弛中值滤波的高阶彩色图像迭代去噪算法[J]. 应用光学, 2016,37(3):366-371.
[14] Lu B B, Li Y, Wang Y M . Color image denoising using high order iterating model by combining relaxed median filter[J]. Journal of Applied Optics, 2016,37(3):366-371.
[15] Donoho D L, Johnstone I M . Ideal spatial adaptation by wavelet shrinkage[J]. Biometrika, 1994,81(3):425-455.
doi: 10.1093/biomet/81.3.425 url: https://academic.oup.com/biomet/article-lookup/doi/10.1093/biomet/81.3.425
[16] 王蓓, 张根耀, 李智 , 等. 基于新阈值函数的小波阈值去噪算法[J]. 计算机应用, 2014,34(5):1499-1502.
doi: 10.11772/j.issn.1001-9081.2014.05.1499 url: http://d.wanfangdata.com.cn/Periodical_jsjyy201405063.aspx
[16] Wang B, Zhang G Y, Li Z , et al. Wavelet threshold denoising algorithm based on new threshold function[J]. Journal of Computer Applications, 2014,34(5):1499-1502.
[17] Zhao R M, Cui H M. Improved threshold denoising method based on wavelet transform [C]//Proceedings of the 7th International Conference on Modelling,Identification and Control.Sousse:IEEE, 2015: 1-4.
[18] 胡然, 郭成城, 杨剑锋 . 基于小波阈值和主成分分析的视频去噪算法[J]. 计算机科学, 2016,43(4):290-293.
doi: 10.11896/j.issn.1002-137X.2016.4.059 url: http://d.wanfangdata.com.cn/Periodical/jsjkx201604059
[18] Hu R, Guo C C, Yang J F . Video denoising algorithm based on wavelet threshold and PCA[J]. Computer Science, 2016,43(4):290-293.
[19] Zhang S, Jing H J.Fast log-gabor-based nonlocal means image denoising methods[C]//Proceedings of 2014 IEEE International Conference on Image Processing.Paris:IEEE, 2014: 2724-2728.
[20] 魏宁, 杨元琴, 董方敏 , 等. 多模图像交叉双域滤波算法[J]. 中国图象图形学报, 2016,21(6):691-697.
doi: 10.11834/jig.20160602 url: http://www.cqvip.com/QK/90287A/201606/669146609.html
[20] Wei N, Yang Y Q, Dong F M , et al. Cross dual-domain filter for denoising multi-mode images[J]. Journal of Image and Graphics, 2016,21(6):691-697.
[21] Lexin A, Nadler B, Durand F, et al. Patch complexity,finite pixel correlations and optimal denoising [C]//Proceedings of the 12th European Conference on Computer Vision.Florence:Springer, 2012: 73-86.
[22] Chen Q, Montesinos P, Sun Q S , et al. Adaptive total variation denoising based on difference curvature[J]. Image and Vision Computing, 2010,28(3):298-306.
doi: 10.1016/j.imavis.2009.04.012 url: http://linkinghub.elsevier.com/retrieve/pii/S0262885609000791
[23] Liu J, Wang Y H, Su K J , et al. Image denoising with multidirectional shrinkage in directionlet domain[J]. Signal Processing, 2016,125:64-78.
doi: 10.1016/j.sigpro.2016.01.013 url: http://linkinghub.elsevier.com/retrieve/pii/S016516841600030X
[24] 张凡 . 基于改进NAS-RIF算法的遥感噪声图像自适应复原[J]. 国土资源遥感, 2015,27(2):105-111.doi: 10.6046/gtzyyg.2015.02.17.
doi: 10.6046/gtzyyg.2015.02.17 url: 年度引用
[24] Zhang F . Self-adaptive restoration for remote sensing noise images based on improved NAS-RIF algorithm[J]. Remote Sensing for Land and Resources, 2015,27(2):105-111.doi: 10.6046/gtzyyg.2015.02.17.
[25] 吴一全, 吴超 . 结合NSCT和KPCA的高光谱遥感图像去噪[J]. 遥感学报, 2012,16(3):533-544.
doi: 10.11834/jrs.20121018 url: http://www.cqvip.com/QK/92457A/201203/41860840.html
[25] Wu Y Q, Wu C . Denoising of hyperspectral remote sensing images using NSCT and KPCA[J]. Journal of Remote Sensing, 2012,16(3):533-544.
[1] ZANG Liri, YANG Shuwen, SHEN Shunfa, XUE Qing, QIN Xiaowei. A registration algorithm of images with special textures coupling a watershed with mathematical morphology[J]. Remote Sensing for Natural Resources, 2022, 34(1): 76-84.
[2] GUO Xiaozheng, YAO Yunjun, JIA Kun, ZHANG Xiaotong, ZHAO Xiang. Information extraction of Mars dunes based on U-Net[J]. Remote Sensing for Natural Resources, 2021, 33(4): 130-135.
[3] LIU Chunting, FENG Quanlong, JIN Dingjian, SHI Tongguang, LIU Jiantao, ZHU Mingshui. Application of random forest and Sentinel-1/2 in the information extraction of impervious layers in Dongying City[J]. Remote Sensing for Natural Resources, 2021, 33(3): 253-261.
[4] WANG Rong, ZHAO Hongli, JIANG Yunzhong, HE Yi, DUAN Hao. Application and analyses of texture features based on GF-1 WFV images in monthly information extraction of crops[J]. Remote Sensing for Natural Resources, 2021, 33(3): 72-79.
[5] GAO Kaixuan, JIAO Haiming, WANG Xinchuang. Inversion model of forest canopy height based on image texture,spectral and topographic features[J]. Remote Sensing for Land & Resources, 2020, 32(3): 63-70.
[6] Hui HUANG, Xiongwei ZHENG, Genyun SUN, Yanling HAO, Aizhu ZHANG, Jun RONG, Hongzhang MA. Seismic image classification based on gravitational self-organizing map[J]. Remote Sensing for Land & Resources, 2019, 31(3): 95-103.
[7] Jun YANG, Jianjie PEI. An improved ICM algorithm for remote sensing image segmentation[J]. Remote Sensing for Land & Resources, 2018, 30(3): 18-25.
[8] Li YAN, Xun GONG, Hong XIE. Phase correlation supported feature track- matching algorithm for repeating texture[J]. Remote Sensing for Land & Resources, 2018, 30(1): 1-6.
[9] SU Tengfei, ZHANG Shengwei, LI Hongyu. Segmentation algorithm based on texture feature and region growing for high-resolution remote sensing image[J]. REMOTE SENSING FOR LAND & RESOURCES, 2017, 29(2): 72-81.
[10] WANG Yu, FU Meichen, WANG Li, WANG Changyao. Tree-cotton intercropping land extraction based on multi-source high resolution satellite imagery[J]. REMOTE SENSING FOR LAND & RESOURCES, 2017, 29(2): 152-159.
[11] HUA Yongchun, LI Zengyuan, GAO Zhihai, GUO Zhong. Extraction of Nitaria tangutorum dune in Minqin County based on GF-2 data[J]. REMOTE SENSING FOR LAND & RESOURCES, 2017, 29(1): 71-77.
[12] HU Hualong, XUE Wu, QIN Zhiyuan. Extraction of residential area from high resolution images based on wavelet texture and primitive merging[J]. REMOTE SENSING FOR LAND & RESOURCES, 2017, 29(1): 21-28.
[13] ZHANG Enbing, QIN Kun, YUE Mengxue, ZHANG Ye, ZENG Cheng. Feature description and extraction of residential area based on variogram function and grid division[J]. REMOTE SENSING FOR LAND & RESOURCES, 2016, 28(4): 149-155.
[14] XIANG Shengwen, WEN Gongjian, GAO Feng. Knowledge driven change detection method for aircraft targets[J]. REMOTE SENSING FOR LAND & RESOURCES, 2016, 28(4): 77-82.
[15] LIU Wei, ZHAO Wenjie, LI Cheng, LI Ting, TAN Haifeng, MA Yangming. Mean shift object tracking based on center symmetric-local similarity number model[J]. REMOTE SENSING FOR LAND & RESOURCES, 2016, 28(3): 37-45.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
京ICP备05055290号-2
Copyright © 2017 Remote Sensing for Natural Resources
Support by Beijing Magtech