Chen P, Cai X W, Zhao D D, et al. Despeckling for side-scan sonar images based on adaptive block-matching and 3D filtering[J]. Opto-Electron Eng, 2020, 47(7): 190580. doi: 10.12086/oee.2020.190580
Citation: Chen P, Cai X W, Zhao D D, et al. Despeckling for side-scan sonar images based on adaptive block-matching and 3D filtering[J]. Opto-Electron Eng, 2020, 47(7): 190580. doi: 10.12086/oee.2020.190580

Despeckling for side-scan sonar images based on adaptive block-matching and 3D filtering

    Fund Project: Supported by the National Key Research and Development Program of China (2016YFC0301604), Sanya City Special Scientific Research Project (2017KS13), and the Fundamental Research Funds for the Provincial Universities of Zhejiang (RF-C2019001)
More Information
  • Side-scan sonar (SSS) is an electronic device that utilizes the propagation characteristics of sound waves under water to complete underwater detection. Because the SSS produces images and maps according to the intensity of acoustic echo, speckle noise will be inevitably involved. A speckle denoising method based on block-matching and 3D filtering (BM3D) is proposed to filter the multiplicative speckle noise in SSS images. First, the SSS image is transformed by power and logarithm. The wavelet transform is used to estimate the general noisy level of the polluted image. Second, the parameters of the BM3D algorithm are updated according to the noise estimation results of each local patch. At last, after comparing the general noise estimation and the local noise estimation, the proposed algorithm chooses the best estimation to filter every patch separately to solve the problem that the noise is not evenly distributed. The experimental results show that the improved BM3D algorithm can effectively reduce the speckle noise in SSS images and obtain good visual effects. The Equivalent Number of Looks of the proposed algorithm is at least 6.83% higher, the Speckle Suppression Index is lower than traditional algorithm, and the Speckle Suppression and Mean Preservation Index is reduced by at least 3.30%. This method is mainly used for sonar image noise reduction, and has certain practical values for ultrasonic, radar or OCT images polluted by speckle noise.
  • 加载中
  • [1] Blondel P. The Handbook of Sidescan Sonar[M]. Berlin: Springer, 2009: 1-6.

    Google Scholar

    [2] Ye X F, Li P, Deng Y Y. A side scan sonar image denoising algorithm based on compound of fuzzy weighted average and Kalman filter[C]//Proceedings of 2012 IEEE International Conference on Mechatronics and Automation, Chengdu, China, 2012: 720-724.

    Google Scholar

    [3] Devapal D, Kumar S S, Sethunadh R. Discontinuity adaptive SAR image despeckling using curvelet-based BM3D technique[J]. International Journal of Wavelets, Multiresolution and Information Processing, 2019, 17(3): 1950016. doi: 10.1142/S0219691319500164

    CrossRef Google Scholar

    [4] Lee J S, Grunes M R, De Grandi G. Polarimetric SAR speckle filtering and its implication for classification[J]. IEEE Transactions on Geoscience and Remote Sensing, 1999, 37(5): 2363-2373. doi: 10.1109/36.789635

    CrossRef Google Scholar

    [5] Kuan D T, Sawchuk A A, Strand T C, et al. Adaptive noise smoothing filter for images with signal-dependent noise[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1985, PAMI-7(2): 165-177. doi: 10.1109/TPAMI.1985.4767641

    CrossRef Google Scholar

    [6] Frost V S, Stiles J A, Shanmugan K S, et al. A model for radar images and its application to adaptive digital filtering of multiplicative noise[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, PAMI-4(2): 157-166.

    Google Scholar

    [7] Buades A, Coll B, Morel J M. Non-local means denoising[J]. Image Processing On Line, 2011, 1: 208-212. doi: 10.5201/ipol.2011.bcm_nlm

    CrossRef Google Scholar

    [8] Wang J, Guo Y W, Ying Y T, et al. Fast non-local algorithm for image denoising[C]//Proceedings of 2006 International Conference on Image Processing, Atlanta, GA, USA, 2006: 1429-1432.

    Google Scholar

    [9] 李世文, 张彬, 刘泽民, 等.基于波原子阈值算法的OCT图像降噪技术[J].光电工程, 2014, 41(7): 75-80. doi: 10.3969/j.issn.1003-501X.2014.07.013

    CrossRef Google Scholar

    Li S W, Zhang B, Liu Z M, et al. Noise reduction for OCT images based on wave-atom thresholding algorithm[J]. Opto-Electronic Engineering, 2014, 41(7): 75-80. doi: 10.3969/j.issn.1003-501X.2014.07.013

    CrossRef Google Scholar

    [10] 王帆, 陈明惠, 高乃珺, 等.基于字典算法的OCT图像散斑稀疏降噪[J].光电工程, 2019, 46(6): 180572. doi: 10.12086/oee.2019.180572

    CrossRef Google Scholar

    Wang F, Chen M H, Gao N J, et al. OCT image speckle sparse noise reduction based on dictionary algorithm[J]. Opto-Electronic Engineering, 2019, 46(6): 180572. doi: 10.12086/oee.2019.180572

    CrossRef Google Scholar

    [11] Lebrun M. An analysis and implementation of the BM3D image denoising method[J]. Image Processing On Line, 2012, 2: 175-213. doi: 10.5201/ipol.2012.l-bm3d

    CrossRef Google Scholar

    [12] Gao J B, Chen Q, Blasch E. Image denoising in the presence of non-Gaussian, power-law noise[C]//Proceedings of 2012 IEEE National Aerospace and Electronics Conference (NAECON), Dayton, OH, USA, 2012: 103-108.

    Google Scholar

    [13] Yang J Y, Zhang X, Yue H J, et al. IBM3D: integer BM3D for efficient image denoising[J]. Circuits, Systems, and Signal Processing, 2019, 38(2): 750-763.

    Google Scholar

    [14] Dabov K, Foi A, Katkovnik V, et al. BM3D image denoising with shape-adaptive principal component analysis[C]//Signal Processing with Adaptive Sparse Structured Representations (SPARS), Saint-Malo, France, 2009: 1-6.

    Google Scholar

    [15] Zhong H, Ma K, Zhou Y. Modified BM3D algorithm for image denoising using nonlocal centralization prior[J]. Signal Processing, 2015, 106: 342-347. doi: 10.1016/j.sigpro.2014.08.014

    CrossRef Google Scholar

    [16] Zhao T T, Hoffman J, McNitt‐Gray M, et al. Ultra‐low‐dose CT image denoising using modified BM3D scheme tailored to data statistics[J]. Medical Physics, 2019, 46(1): 190-198. doi: 10.1002/mp.13252

    CrossRef Google Scholar

    [17] Cheng W M, Zhu X, Chen X, et al. Manhattan distance-based adaptive 3D transform-domain collaborative filtering for laser speckle imaging of blood flow[J]. IEEE Transactions on Medical Imaging, 2019, 38(7): 1726-1735. doi: 10.1109/TMI.2019.2896007

    CrossRef Google Scholar

    [18] 范习健, 李庆武, 黄河, 等.侧扫声呐图像的3维块匹配降斑方法[J].中国图象图形学报, 2012, 17(1): 68-74.

    Google Scholar

    Fan X J, Li Q W, Huang H, et al. Side-scan sonar image despeckling based on block-matching and 3D filtering[J]. Journal of Image and Graphics, 2012, 17(1): 68-74.

    Google Scholar

    [19] Dabov K, Foi A, Katkovnik V, et al. Image denoising by sparse 3-D transform-domain collaborative filtering[J]. IEEE Transactions on Image Processing, 2007, 16(8): 2080-2095. doi: 10.1109/TIP.2007.901238

    CrossRef Google Scholar

    [20] James R, Supriya M H. Blind estimation of single look side scan sonar image from the observation model[J]. Procedia Computer Science, 2016, 93: 336-343. doi: 10.1016/j.procs.2016.07.218

    CrossRef Google Scholar

    [21] 张洪科. 数字全息再现像中的噪声抑制[D]. 哈尔滨工程大学,2014. http://cdmd.cnki.com.cn/Article/CDMD-10217-1017240756.htm.

    Google Scholar

    Zhang H K. Suppression of Noise in Digital Holography[D]. Harbin Engineering University, 2014. http://cdmd.cnki.com.cn/Article/CDMD-10217-1017240756.htm.

    Google Scholar

  • Overview: An understanding of the ocean and its changing environment is increasingly important. Scientific, economic, and political decision-making depends to some extent on this knowledge. However, even lasers can penetrate through only a few tens of meters in very clear water. Acoustic waves, by contrast, can travel over long distances without much attenuation. Therefore, all kinds of sonars play an important role in ocean research. Side-scan sonar (SSS) is an electronic device that utilizes the propagation characteristics of sound waves under water to complete underwater detection and communication tasks through electro-acoustic conversion and information processing. Because the SSS produces images and maps according to the intensity of acoustic echo, speckle noise will inevitably be involved due to the complex underwater environments. Block-matching and 3D filtering (BM3D) is an advanced denoising method based on the fact that an image has a locally sparse representation in transform domain. This sparsity is enhanced by grouping similar 2D image patches into 3D groups. This algorithm performs well in dealing images polluted by Gaussian additive noise. The BM3D algorithm was originally designed for Gaussian additive noise, therefore, it is not reasonable to denoise the side-scan sonar images polluted by speckle noise. In this paper, a speckle denoising method based on BM3D is proposed to filter the multiplicative speckle noise in side-scan sonar images. First, the SSS image is transformed by power and logarithm. The multi-scale two-dimensional discrete wavelet transform is used to estimate the general noisy level of the polluted image. Second, the parameters of the BM3D algorithm are updated according to the noise estimation results of each local patch. Third, after comparing the general noise estimation and the local noise estimation, the proposed algorithm chooses the best estimation to filter every patch separately to solve the problem that the noise is not evenly distributed. Finally, the image properties are recovered by exponential transformation and inverse power transformation. The experimental results show that the improved BM3D algorithm can effectively reduce the speckle noise in SSS images and obtain good visual effects. In this paper, three non-reference image quality evaluation parameters, namely the equivalent noise of looks (ENL), speckle reduction index (SSI), speckle suppression and average preservation index (SMPI), are used to evaluate the noise reduction effect. Compared with two kinds of improved BM3D algorithms and a traditional algorithm, the ENL of the proposed algorithm is at least 6.83% higher than that of others, its SSI is very similar to that of Manhattan distance-based adaptive block-matching and 3D filtering(MD-ABM3D), and its SMPI is reduced by at least 3.30%. This method is mainly used for sonar image noise reduction, and has certain practical values for ultrasonic, radar or OCT images polluted by speckle noise.

  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Figures(4)

Tables(2)

Article Metrics

Article views(7260) PDF downloads(2693) Cited by(0)

Access History
Article Contents

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint