基于非局部自相似序列集的一类视频图像盲去噪算法

doi:10.12263/DZXB.20191207

电子学报 ›› 2021, Vol. 49 ›› Issue (8): 1498-1506.DOI: 10.12263/DZXB.20191207

基于非局部自相似序列集的一类视频图像盲去噪算法

邢远秀¹^,², 李军贤¹^,², 王文波¹^,², 王松³

^1.武汉科技大学理学院，湖北武汉 430081
^2.冶金工业过程系统科学湖北省重点实验室，湖北武汉 430081
^3.南卡罗莱纳大学工程与计算学院，美国南卡罗莱纳州哥伦比亚市 29208

收稿日期:2019-10-25 修回日期:2021-02-24 出版日期:2021-08-25
- 作者简介:
- 邢远秀（通信作者）女，1980年8月出生于河南省南阳市.博士，武汉科技大学理学院副教授，硕士生导师.主要研究方向为图像处理、机器视觉等. E-mail：yuanxiu@126.com
  李军贤女，1994年1月出生于河南省安阳市.现为武汉科技大学硕士研究生，主要研究方向为图像处理、计算机视觉.E-mail：1156325080@qq.com
- 基金资助:
- 冶金工业过程系统科学湖北省重点实验室基金 (Y202008); 国家自然科学基金 (61671338)

Blind Video Image Denoising Based on Nonlocal Self-Similarity Series Sets

XING Yuan-xiu¹^,², LI Jun-xian¹^,², WANG Wen-bo¹^,², WANG Song³

^1.College of Science, Wuhan University of Science and Technology, Wuhan, Hubei 430081, China
^2.Hubei Province Key Laboratory of Systems Science in Metallurgical Process, Wuhan, Hubei 430081, China
^3.College of Engineering and Computing, University of South Carolina, Columbia, SC 29208, USA

Received:2019-10-25 Revised:2021-02-24 Online:2021-08-25 Published:2021-08-25
- Supported by:
- Fund for Hubei Province Key Laboratory of Systems Science in Metallurgical Process (Y202008); National Natural Science Foundation of China (61671338)

摘要/Abstract

摘要：

为提高具有帧间位移平移特性的视频图像的信噪比和去噪时效，本文提出了基于非局部自相似序列集的视频图像盲去噪算法.选取与待去噪视频图像前后相邻的若干图像帧，在每一图像帧中寻找具有典型特征的图像块群，并通过在前一帧图像中查找和该图像块群具有最小差异度的块群来确定帧间的精确位移；将待去噪视频图像划分成若干图像块，根据帧间位移快速构建每个图像块的自相似序列集；随后将每个自相似序列集中的二维图像块整合成三维矩阵后进行三维变换，并对变换系数进行自适应阈值处理；再将三维逆变换后的图像块融合生成去噪图像.实验结果表明，在噪声方差未知的情况下，本文算法所得去噪视频图像具有较好的信噪比和视觉效果，并且有较高的运行效率.

关键词: 视频图像去噪, 块匹配, 自相似序列集, 三维变换, 自适应阈值

Abstract:

A blind video image denoising algorithm based on non-local self-similar series sets is proposed to improve the peak signal-to-noise ratio and denoising efficiency of video images with displacement characteristics. Image block-groups with typical features are detected in each frame, and accurate inter-frame displacement between the image and previous frame is computed based on image block-matching. The noise image is divided into several image blocks, and the self-similar series set of each image blocks is constructed quickly according to the inter-frame displacement. A 3D transform is applied to the self-similar series set, followed by an adaptive-threshold of the transform coefficients to attenuate the noise. The 3D estimate after inverse 3D transformation are aggregated to obtain the finial image. Experimental results show that the proposed algorithm has significate advantages in PSNR and visual effect in unknown noise variance, meanwhile it also has higher efficiency.

Key words: video image denoising, block-matching, self-similarity series set, 3D transform, adaptive threshold

中图分类号:

TP391

邢远秀, 李军贤, 王文波, 等. 基于非局部自相似序列集的一类视频图像盲去噪算法[J]. 电子学报, 2021, 49(8): 1498-1506.

Yuan-xiu XING, Jun-xian LI, Wen-bo WANG, et al. Blind Video Image Denoising Based on Nonlocal Self-Similarity Series Sets[J]. Acta Electronica Sinica, 2021, 49(8): 1498-1506.

图/表 8

图1 块群结构示例（m为块群中块的个数）

图2 测试图像

图3 9种算法在高斯噪声方差σ=30的人物图像1上的去噪效果

图4 9种算法在高斯噪声方差σ=20的人物图像2上的去噪效果

图5 9种算法在不同高斯方差下去噪后的平均信噪比

表1 9种不同算法在人物图像上所需平均处理时间 (s)

	NL-Means	FNL-Means	BM3D	WNNM	SNSS	GSRC-NLP	SALT	E-RF3D	本文算法
人物图像1	70.068	2.347	124.503	327.614	977.470	212.411	334.612	4.011	9.1684
人物图像2	67.757	2.228	121.449	345.235	946.561	203.144	301.087	4.039	9.0511

图6 9种算法在井筒视频图像帧上的去噪效果

表2 9种算法在井筒视频图像帧上所需平均处理时间 (s)

	NL-Means	FNL-Means	BM3D	WNMM	SNSS	GSRC-NLP	SALT	E-RF3D	本文算法
时间	20.065	0.397	38.775	88.217	187.384	54.221	76.862	1.5756	2.7837

参考文献 29

1	DodgeS, KaramL. Understanding how image quality affects deep neural networks[A]. 2016 Eighth International Conference on Quality of Multimedia Experience[C]. Lisbon, Portugal: IEEE, 2016. 1－6.
2	HeK M, ZhangX Y, RenS Q, et al. Deep residual learning for image recognition[A]. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)[C]. Las Vegas, USA: IEEE, 2016. 770－778.
3	ChenJ W, ChenJ, ChaoH Y, et al. Image blind denoising with generative adversarial network based noise modeling[A]. 2018 IEEE Conference on Computer Vision and Pattern Recognition[C]. Salt Lake City, USA: IEEE, 2018. 3155－3164.
4	FoiA, KatkovnikV, EgiazarianK. Pointwise shape-adaptive DCT for high-quality denoising and deblocking of grayscale and color images[J]. IEEE Transactions on Image Processing, 2007, 16(5): 1395－1411.
5	YangJ H, FengR, DengW. A new algorithm of image denoising based on stationary wavelet multi-scale adaptive threshold[A]. Proceedings of 2011 International Conference on Electronic & Mechanical Engineering and Information Technology[C]. Harbin, China: IEEE, 2011. 4550－4553.
6	YangJ, WrightJ, HuangT S, et al. Image super-resolution via sparse representation[J]. IEEE Transactions on Image Processing, 2010, 19(11): 2861－2873.
7	WenB, RavishankarS, BreslerY. Structured overcomplete sparsifying transform learning with convergence guarantees and applications[J]. International Journal of Computer Vision, 2015, 114(2/3): 137－167.
8	BuadesA, CollB, MorelJ M. A non-local algorithm for image denoising[A]. 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)[C]. San Diego, USA: IEEE, 2005. 60－65.
9	FromentJ. Parameter-free fast pixelwise non-local means denoising[J]. Image Processing on Line, 2014, 4: 300－326.
10	FrosioI, KautzJ. Statistical nearest neighbors for image denoising[J]. IEEE Transactions on Image Processing, 2019, 28(2): 723－738.
11	WangS L, ZhangL, LiangY. Nonlocal spectral prior model for low-level vision[A]. Computer Vision - ACCV[M]. Berlin, Heidelberg, GER: Springer, 2013. 231－244.
12	GuS H, XieQ, MengD Y, et al. Weighted nuclear norm minimization and its applications to low level vision[J]. International Journal of Computer Vision, 2017, 121(2): 183－208.
13	宋云, 李雪玉, 沈燕飞, 等. 基于非局部相似块低秩的压缩感知图像重建算法[J]. 电子学报, 2017, 45(3): 695－703.
	SongY, LiX Y, ShenY F, et al. Compressed sensing image reconstruction based on low rank of non-local similar patches[J]. Acta Electronica Sinica, 2017, 45(3): 695－703. (in Chinese)
14	ZhaZ Y, YuanX, WenB H, et al. From rank estimation to rank approximation: Rank residual constraint for image restoration[J]. IEEE Transactions on Image Processing, 2020, 29: 3254－3269.
15	WenB H, LiY J, BreslerY. Image recovery via transform learning and low-rank modeling: The power of complementary regularizers[J]. IEEE Transactions on Image Processing, 2020, 29: 5310－5323.
16	DabovK, FoiA, KatkovnikV, et al. Image denoising by sparse 3-D transform-domain collaborative filtering[J]. IEEE Transactions on Image Processing, 2007, 16(8): 2080－2095.
17	DjurovićI. BM3D filter in salt-and-pepper noise removal[J]. EURASIP Journal on Image and Video Processing, 2016, 10(1): 1－11.
18	LiY J, ZhangJ W, WangM N. Improved BM3D denoising method[J]. IET Image Processing, 2017, 11(12): 1197－1204.
19	YangJ Y, ZhangX, YueH J, et al. IBM3D: Integer BM3D for efficient image denoising[J]. Circuits, Systems, and Signal Processing, 2019, 38(2): 750－763.
20	ZhaZ Y, YuanX, ZhouJ T, et al. Image restoration via simultaneous nonlocal self-similarity priors[J]. IEEE Transactions on Image Processing, 2020, 29: 8561－8576.
21	ZhaZ Y, YuanX, WenB H, et al. Group sparsity residual constraint with non-local priors for image restoration[J]. IEEE Transactions on Image Processing, 2020, 29: 8960－8975.
22	BuadesA, LisaniJ L. Enhancement of noisy and compressed videos by optical flow and non-local denoising[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2020, 30(7): 1960－1974.
23	LiuC, FreemanW T. A high-quality video denoising algorithm based on reliable motion estimation[A]. European Conference on Computer Vision[C]. Berlin, Germany: Springer, 2010. 706－719.
24	WenB H, LiY J, PfisterL, et al. Joint adaptive sparsity and low-rankness on the fly: An online tensor reconstruction scheme for video denoising[A]. 2017 IEEE International Conference on Computer Vision[C]. Venice, Italy: IEEE, 2017. 241－250.
25	DabovK, FoiA, EgiazarianK. Video denoising by sparse 3D transform-domain collaborative filtering[A]. 2007 15th European Signal Processing Conference[C]. Poznań, Poland: IEEE, 2010. 145－149.
26	肖进胜, 姜红, 彭红, 等. 一种改进的3维块匹配视频去噪算法[J]. 四川大学学报(工程科学版), 2014, 46(4): 81－86.
	XiaoJ S, JiangH, PengH, et al. An improved video denoising algorithm based on 3D block matching[J]. Journal of Sichuan University (Engineering Science Edition), 2014, 46(4): 81－86. (in Chinese)
27	MaggioniM, BoracchiG, FoiA, et al. Video denoising, deblocking, and enhancement through separable 4-D nonlocal spatiotemporal transforms[J]. IEEE Transactions on Image Processing, 2012, 21(9): 3952－3966.
28	MaggioniM, Sánchez-MongeE, FoiA. Joint removal of random and fixed-pattern noise through spatiotemporal video filtering[J]. IEEE Transactions on Image Processing, 2014, 23(10): 4282－4296.
29	邢远秀. 矿山井筒视觉监测与故障智能识别系统设计[J]. 金属矿山, 2013, (9): 138－141.
	XingY X. Design of the vision monitoring and fault intelligent recognition system for security of mineshaft[J]. Metal Mine, 2013, (9): 138－141. (in Chinese)

[1]	陈震, 倪晶晶, 张聪炫, 葛利跃, 王梓歌. 基于金字塔块匹配的双目场景流估计[J]. 电子学报, 2022, 50(9): 2164-2171.
[2]	张聪炫, 陈震, 熊帆, 黎明, 葛利跃, 陈昊. 非刚性稠密匹配大位移运动光流估计[J]. 电子学报, 2019, 47(6): 1316-1323.
[3]	凌明, 武建平, 冯克环. 一种Dalvik虚拟机的自适应编译系统[J]. 电子学报, 2013, 41(8): 1622-1627.
[4]	王蓉芳, 焦李成, 刘芳, 杨淑媛. 利用纹理信息的图像分块自适应压缩感知[J]. 电子学报, 2013, 41(8): 1506-1514.
[5]	王颖;李洁;高新波. 基于MCA的乳腺X线图像中肿块的自适应检测方法[J]. 电子学报, 2011, 39(3): 525-530.
[6]	梁栋, 殷兵, 于梅, 李新华, 王年, . 基于非抽样Contourlet变换的自适应阈值图像增强算法[J]. 电子学报, 2008, 36(3): 527-530.
[7]	宋建斌, 李波, 李炜, 吴波. 适用于H.264/AVC的快速帧内预测算法[J]. 电子学报, 2007, 35(4): 668-672.
[8]	段哲民, 周巍, 李辉, 周欣. 一种基于H.264的有效运动估计算法[J]. 电子学报, 2006, 34(S1): 2406-2409.
[9]	孙慧平, 刘党辉, 沈兰荪. 基于DCT压缩域的快速字符定位算法研究[J]. 电子学报, 2006, 34(4): 751-754.
[10]	魏伟, 侯正信, 郭迎春. 基于节点搜索的可变形块运动补偿[J]. 电子学报, 2005, 33(8): 1421-1424.
[11]	何卫锋;毛志刚. 基于帧级流水脉动阵列结构的运动估计电路[J]. 电子学报, 2005, 33(8): 1487-1491.
[12]	鲁照华;李华;褚晶辉;俞斯乐. 一种预测质量可控的快速运动估计搜索算法[J]. 电子学报, 2004, 32(6): 980-982.
[13]	安平;张兆扬;马然. 应用分层MRF/GRF模型的立体图像视差估计及分割[J]. 电子学报, 2003, 31(4): 597-601.
[14]	李玙;徐维朴;郑南宁;王辉. 一种新的基于遗传算法的快速运动估计方法[J]. 电子学报, 2000, 28(6): 114-117.

基于非局部自相似序列集的一类视频图像盲去噪算法

Blind Video Image Denoising Based on Nonlocal Self-Similarity Series Sets

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 29

相关文章 14

编辑推荐

Metrics