基于边缘检测和聚类分析的浅剖图像分层算法
A New Method of Classification on Shallow Sub-Bottom Profiles Based on Image Edge Detection and Cluster Analysis
DOI: 10.12677/JISP.2019.82008, PDF,   
作者: 徐迎晨*：江苏省张家港市长江治理工程管理处，江苏 张家港；孙佳龙：北京大地宏图勘测科技有限公司，北京
关键词: 边缘检测；聚类分析；浅剖图像；地层识别；Edge Detection； Cluster Analysis； Shallow Sub-Bottom Profiles； Layer Recognition
摘要: 本文提出了一种基于边缘检测和聚类分析的浅剖图像分层算法。利用自适应的Canny边缘检测算法有效提取了图像边缘，利用区域跟踪算法，确定了二值图像中所有目标的边界，根据k-means聚类分析算法，对多条边界进行了有效分类，实现了地层划分。以钻孔数据为参考，比较了基于边缘检测和聚类分析的浅剖图像分层算法与ISE软件处理结果的精度。结果表明，本文提出的方法在海底高程和三个层底高程的识别中，中误差为0.47米，远小于ISE的1.37米，说明该方法优于ISE处理的结果，可以为处理浅地层剖面图像提供一个新的思路和方法。
Abstract: A new method of classification on shallow sub-bottom profiles based on edge detection and cluster analysis is presented in the paper. The sub-bottom profile image edge is effectively extracted by Canny edge detection algorithm, and image boundaries of all objects in binary image are deter-mined by region tracking algorithm. According to the k-means clustering algorithm, effective clas-sification of boundary and the stratigraphic division are realized in this paper. Taking drill hole data for reference, the results from the paper’s algorithm and ISE software are compared. The re-sults show that the error of mean square is 0.47 meters according to the proposed method, far less than 1.37 meters of ISE. Therefore the method is superior to the ISE processing results and can provide a new idea and method for the treatment of shallow profile image.
文章引用：徐迎晨, 孙佳龙. 基于边缘检测和聚类分析的浅剖图像分层算法[J]. 图像与信号处理, 2019, 8(2): 51-59. https://doi.org/10.12677/JISP.2019.82008

参考文献

[1] Wang, F., Qi, F., Hu, G., et al. (2012) Correction of Seabed Layer Thickness in Processing Subbottom Profile Data. Marine Science Bulletin, 14, 83-96.
[2] 金翔龙. 海洋地球物理研究与海底探测声学技术的发展[J]. 地球物理学进展, 2007, 22(4): 1243-1249.
[3] 赵荻能, 吴自银, 周洁琼, 等. 声速剖面精简运算的改进D-P算法及其评估[J]. 测绘学报, 2014, 43(7): 681-689.
[4] 沈远海, 马远良, 屠庆平, 等. 浅水声速剖面的反演方法与实验验证[J]. 西北工业大学学报, 2000, 18(2): 212-215.
[5] 罗进华, 潘国富, 丁维风. 消除涌浪对海底声学地层剖面影响的处理技术研究[J]. 声学技术, 2009, 28(1): 21-24.
[6] 王方旗, 亓发庆, 姚菁, 等. 浅海区C-Boom型浅地层剖面地层畸变及校正[J]. 海洋科学进展, 2011, 29(1): 47-53.
[7] 李一保, 张玉芬, 刘玉兰, 等. 浅地层剖面仪在海洋工程中的应用[J]. 工程地球物理学报, 2007, 4(1): 4-8.
[8] 丁维凤, 罗进华, 来向华, 等. 浅地层剖面交互拾取解释技术研究[J]. 海洋科学, 2008, 32(9): 1-6.
[9] Lin, Y.T., Schuettpelz, C.C., Wu, C.H., et al. (2009) A Combined Acoustic and Electromagnetic Wave-Based Techniques for Bathymetry and Subbottom Profiling in Shallow Waters. Journal of Applied Geophysics, 68, 203-218.
[Google Scholar] [CrossRef] 
[10] Bulla, J.M., Gutowskia, M., Dixa, J.K., Henstocka, T.J. and Hogarth, P. (2005) Design of 3D Chirp Sub-Bottom Imaging System. Marine Geophysical Researches, 26, 157-169.
[11] Deimling, J.S.V., Held, P., Feldens, P., et al. (2015) Effects of Using Inclined Parametric Echosounding on Sub-Bottom Acoustic Imaging and Advances in Buried Object Detection. Geo-Marine Letters, 1-7.
[12] Ha, H.K., Maa, P.Y., Park, K., et al. (2011) Estimation of High-Resolution Sediment Concentration Profiles in Bottom Boundary Layer Using Pulse-Coherent Acoustic Doppler Current Profilers. Marine Ge-ology, 279, 199-209.
[13] Cukur, D., Krastel, S., Çağatay, M.N., et al. (2015) Evidence of Extensive Carbonate Mounds and Subla-custrine Channels in Shallow Waters of Lake Van, Eastern Turkey, Based on High-Resolution Chirp Subbottom Profiler and Multibeam Echosounder Data. Geo-Marine Letters, 35, 1-12.
[14] 于合龙, 刘浩洋, 苏恒强, 等. 基于图像Canny边缘检测和Hough变换算法的高温结构测量方法[J]. 吉林大学学报: 理学版, 2014, 52(3): 519-524.
[15] 韩涛, 祝跃飞. 基于Canny边缘检测的自适应空域隐写术[J]. 电子与信息学报, 2015, 37(5): 1266-1270.
[16] 孙智鹏, 邵仙鹤, 王翥, 等. 改进的自适应Canny边缘检测算法[J]. 电测与仪表, 2016, 53(6): 17-21.
[17] 凌凤彩, 康牧, 林晓. 改进的Canny边缘检测算法[J]. 计算机科学, 2016(8).
[18] 熊忠阳, 陈若田, 张玉芳. 一种有效的K-means聚类中心初始化方法[J]. 计算机应用研究, 2011, 28(11): 4188-4190.
[19] 夏士雄, 李文超, 周勇, 等. 一种改进的k-means聚类算法[J]. 东南大学学报(英文版), 2007, 23(3): 435-438.
[20] 付宁, 乔立岩, 彭喜元. 基于改进K-means聚类和霍夫变换的稀疏源混合矩阵盲估计算法[J]. 电子学报, 2009, 37(s1): 92-96.