FCM-CV水平集算法在沉底小目标声呐图像分割中的应用

韩婷婷; 王璐瑶; 周 天; 徐 超; 张丽红; 李海森

doi:10.11993/j.issn.2096-3920.2021.03.005

FCM-CV水平集算法在沉底小目标声呐图像分割中的应用

doi: 10.11993/j.issn.2096-3920.2021.03.005

1.
哈尔滨工程大学水声技术重点试验室, 黑龙江哈尔滨, 150001
2.
工业和信息化部海洋信息获取与安全工信部重点实验室哈尔滨工程大学, 黑龙江哈尔滨, 150001
3.
哈尔滨工程大学水声工程学院, 黑龙江哈尔滨, 150001

基金项目: 国家重点研发计划项目(2018YFF01013401); 国家自然科学基金(U1709203, 41976176, U1906218); 中央高校基本科研业务费专项资金(3072020CFT0501); 黑龙江省博士后科研发展基金(LBH-Q18042); 黑龙江省自然科学基金项目(ZD 2020D001).

详细信息

作者简介:
韩婷婷(1994-), 女, 硕士, 主要研究方向为声呐图像处理.

中图分类号: TJ630 TN911.7
计量
- 文章访问数: 133
- HTML全文浏览量: 1
- PDF下载量: 126
- 被引次数: 0
出版历程
- 收稿日期: 2020-04-22
- 修回日期: 2020-09-28
- 刊出日期: 2021-06-30

Application of FCM-CV Level Set Algorithm in Sonar Image Segmentation of Small Sinking Target

1.
Acoustic Science and Technology Laboratory, Harbin Engineering University, Harbin 150001, China
2.
Key Laboratory of Marine Information Acquisition and SecurityHarbin Engineering University, Harbin 150001, China
3.
College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China

摘要

摘要: 针对沉底小目标高频声呐图像信混比低, 难以与背景分离的问题, 提出模糊聚类(FCM)算法与Chan-Vese(CV)水平集算法相结合的分割方法。该方法利用FCM算法得到的隶属度函数自动设定水平集模型曲线的初始演化位置, 解决了CV水平集分割因初始位置设置不当而无法得到准确分割结果的不足; 同时根据模糊聚类的结果, 对水平集演化的控制参数进行估计, 使得分割过程更加稳健。通过仿真数据及外场试验数据处理可知, 相较于FCM和马尔科夫随机场分割算法, 文中算法对斑点噪声不敏感, 可分割出完整的边界; 相较于常规CV水平集算法, 文中算法因粗分割后零水平集的设定更接近目标的边缘, 可以在较少迭代次数下即可获得更加精确的分割结果。
- 沉底小目标 /
- 模糊聚类算法 /
- CV水平集分割 /
- 声呐图像 /
- 马尔科夫随机场分割
Abstract: Aiming at the problem that the signal-to-mix ratio of the small sinking target is low, and it is difficult to separate from the background, this study proposes a segmentation method to use the fuzzy c-means(FCM) algorithm to coo- perate with the Chan-Vese(CV) level set. This method automatically sets the initial evolution position of the level set model curve by using the membership function obtained from the FCM algorithm, which solves the problem that the CV-level set segmentation cannot obtain accurate segmentation results because of the incorrect initial position setting. Simultaneously, the control parameters of the level set evolution are estimated according to the results of fuzzy clustering, which makes the segmentation process more robust. Using outfield test and simulation data, compared with the algorithm of FCM segmentation and Markov random field segmentation, the algorithm in this study is not sensitive to speckle noise and cansegmenta complete boundary. Compared with the conventional CV-level set algorithm, the algorithm in this study can obtain more accurate segmentation results with feweriterations.
- small sinking target /
- fuzzy c-means(FCM) algorithm /
- Chan-Vese(CV) level set segmentation /
- sonar image /
- Markov random field segmentation

HTML全文

参考文献(12)

[1]	席志红, 赵春梅. 基于变分水平集理论的水下图像分割方法[J]. 应用科技, 2019, 46(2): 57-62. Xi Zhi-hong, Zhao Chun-mei. An Underwater Image Segmentation Method based on the Variational Level Set Theory[J]. Applied Science and Technology, 2019, 46(2): 57-62.
[2]	王雷, 叶秀芬, 王天. 模糊聚类的侧扫声纳图像分割算法[J]. 华中科技大学学报(自然科学版), 2012, 40(9): 25- 29. Wang Lei, Ye Xiu-fen, Wang Tian. Segmentation Algorithm of Fuzzy Clustering on Side Scan Sonar Image[J]. Journal of Huazhong University of Science and Technology Nature Science, 2012, 40(9): 25-29.
[3]	王雷. 侧扫声呐图像分割算法研究[D]. 哈尔滨: 哈尔滨工程大学, 2013.
[4]	Huang Y, Huang Y C. Segmenting SAR Satellite ImagesWith the Multilayer Level Set Approach[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2011, 4(3): 632-642.
[5]	Mumford D, Shah J. Optimal Approximations by Piecewise Smooth Functions and Associated Variational Problems[J]. Communications on Pure & Applied Mathematics, 1989, 42(5): 577-685.
[6]	Chan T E, Vese L A. A Level Set Algorithm for Minimizing the Mumford-Shah Functional in Image Processing [C]//IEEE Workshop on Variational & Level Set Methods. Vancouver, BC, Canada: IEEE, 2001.
[7]	Li C, Xu C, Gui C, et al. Level Set Evolution without Reinitialization: A New Variational Formulation[C]//IEEE Computer Society Conference on Computer Vision & Pattern Recognition. San Diego, CA, USA: IEEE, 2005.
[8]	Vikas A, Sreenivasan A, Sudha K L. Fuzzy C-means Clustering and Criminisi Algorithm Based Shadow Re-moval Scheme for Side Scan Sonar Images[C]//2017 International Conference on Innovative Mechanisms for Industry Applications(ICIMIA). Bangalore: IEEE, 2017: 411- 414.
[9]	王兴梅. 水下声呐图像的MRF目标检测与水平集的轮廓提取方法研究[D]. 哈尔滨: 哈尔滨工程大学, 2010.
[10]	刘璐. 基于类内类间距离的模糊聚类算法及图像分割应用[D]. 西安: 西安邮电大学, 2016.
[11]	叶秀芬, 王兴梅, 方超, 等. 基于改进的马尔可夫随机场声呐图像分割方法研究[J]. 兵工学报, 2009, 30(8): 1039-1045. Ye Xiu-fen, Wang Xing-mei, Fang Chao, et al. Study of Sonar Imagery Segmentation Algorithm Based on Im-proved Markov Random Field Model[J]. Acta Armamen-tarII, 2009, 30(8): 1039-1045.
[12]	Von K T. The Impact of Seaplane Floats during Landing[R]. Washington, USA: National Advisory Committee for Aeronautics, NACA TN 321, 1929: 1-8.