多音束測深儀(Multibeam Echosounder, MBES)水深測量因其效率以及品質而成為目前水深測量的主流方式，而多音束測深資料中的異常值會對後續製作水下數值高程模型(Digital Elevation Model, DEM)造成極大影響，可能會影響航行安全，因此異常值的偵測及刪除是測深資料處理中重要的步驟。傳統上使用商用軟體手動偵測異常值的方式費時耗工，半自動演算法則依賴使用者經驗設定參數，無法確保過濾資料品質。本研究以台灣多個港口的MBES測深點雲為研究資料，提出以核點全卷積神經網路(Kernel Point Fully Convolutional Neural Network)之深度學習模型應用於測深資料異常值的偵測任務上，應用仿射轉換(Affine Transformation)與實例採樣 (GT sampling)做為資料強化，以點雲三維坐標及以M-估計量(M-estimator)所計算殘差做為訓練特徵，並採用焦點損失函數(Focal Loss)為損失計算方式以增強訓練成效。最後偵測結果與CARIS&HIP軟體人工偵測結果相比較，竹圍漁港異常值偵測結果之正確率(Accuracy)為0.999、精確率(Precision)為0.991、召回率(Recall)為0.735、F1得分(F1 Score)為0.844，水下數值高程模型(Digital Elevation Model, DEM)網格較差之平均值為-0.001公尺和標準差為0.046公尺；碧砂漁港點雲異常值偵測結果之正確率0.973、精確率0.942、召回率0.826、F1 得分0.880，水下DEM網格較差之平均值為0.003公尺和標準差為0.135公尺。模型主要會依據空間位置分布、群集形狀特徵與局部點密度作為偵測依據，因此對於離散的異常值較有可能被偵測，當異常值群集在幾何與密度特徵上與真實地形或結構物高度相似時，模型偵測能力會受到限制。

Multibeam Echosounder (MBES) bathymetric surveys are the preferred method for efficient, high-quality depth mapping, but outliers in the data can compromise digital bathymetric model (DBM) accuracy and navigation safety. Manual classification is labor-intensive, and semi-automatic methods depend on user-defined parameters that may not ensure data quality. This study introduces a deep learning model based on the Kernel Point Fully Convolutional Neural Network (KP-FCNN) to classify outliers in MBES point cloud data from multiple Taiwanese ports. Data augmentation used affine transformation and Ground Truth (GT) sampling; training features included 3D coordinates and M-estimator residuals, with Focal Loss as the loss function. Predictions were compared to manual classification from CARIS HIPS. For Zhuwei Fishing Port, classification accuracy was 0.999, precision 0.991, recall 0.735, and F1 score 0.844; the underwater Digital Elevation Model (DEM) had a mean error of 0.001 m and standard deviation of 0.046 m. For Bisha Fishing Port, accuracy was 0.973, precision 0.942, recall 0.826, and F1 score 0.880; the DEM showed a mean error of 0.003 m and standard deviation of 0.135 m. The model primarily relies on spatial distribution, cluster morphology, and local point density, performing well on isolated outliers but less effectively when outlier clusters resemble genuine terrain or structures.

史天元, & 薛憲文. (2022). 高程現代化於海洋測量之實踐 [The Realization of Height Modernization in Hydrographic Surveying]. 國土測繪與空間資訊, 10(1), 21–36. https://www.airitilibrary.com/Article/Detail?DocID=P20130301007-202201-202201170007-202201170007-21-36
劉庭瑜（2024）。半自動雜訊標記應用於多音束無人船現代化水深測量點雲。﹝碩士論文。國立成功大學﹞臺灣博碩士論文知識加值系統。 https://hdl.handle.net/11296/7ysc8g。
Arge, L., Larsen, K. G., Mølhave, T., & van Walderveen, F. (2010). Cleaning massive sonar point clouds. Proceedings of the 18th SIGSPATIAL International Conference on Advances in Geographic Information Systems, 152–161. https://doi.org/10.1145/1869790.1869815
Bisquay, H., Freulon, X., De Fouquet, C., & Lajaunie, C. (1998). Multibeam data cleaning for hydrography using geostatistics. IEEE Oceanic Engineering Society. OCEANS'98. Conference Proceedings (Cat. No. 98CH36259), 2, 1135-1143. https://doi.org/10.1109/OCEANS.1998.724413
Bourillet, J. F., Edy, C., Rambert, F., Satra, C., & Loubrieu, B. (1996). Swath mapping system processing: Bathymetry and cartography. Marine Geophysical Researches, 18, 487–506. https://doi.org/10.1007/BF00286091
Burke, R., Forbes, S., & White, K. (1988). Processing'Large'Data Sets From 100% Bottom Coverage'Shallow'Water Sweep Surveys A New Challenge for the Canadian Hydrographic Service. The International Hydrographic Review.
Calder, B. R., & Mayer, L. A. (2003). Automatic processing of high-rate, high-density multibeam echosounder data. Geochemistry Geophysics Geosystems, 4(3), 487-506. https://doi.org/10.1029/2002GC000486
Calder, B. R., & Rice, G. (2011). Design and implementation of an extensible variable resolution bathymetric estimator, U.S. HYDRO. CONF. https://doi.org/https://scholars.unh.edu/ccom/841/
Canepa, G., Bergem, O., & Pace, N. G. (2003). A new algorithm for automatic processing of bathymetric data. IEEE Journal of oceanic engineering, 28(1), 62–77. https://doi.org/10.1109/JOE.2002.808204
Chadwell, C. D., & Sweeney, A. D. (2010). Acoustic Ray-Trace Equations for Seafloor Geodesy. Marine Geodesy, 33(2-3), 164–186. https://doi.org/10.1080/01490419.2010.492283
Debese, N., & Bisquay, H. (1999). Automatic detection of punctual errors in multibeam data using a robust estimator. International Hydrographic Review, 76(1), 49–63. https://doi.org/https://journals.lib.unb.ca/index.php/ihr/article/view/26159
Debese, N., Moitié, R., & Seube, N. (2012). Multibeam echosounder data cleaning through a hierarchic adaptive and robust local surfacing. Computers & Geosciences, 46, 330–339. https://doi.org/10.1016/j.cageo.2012.01.012
Devote, B. D. (2024). Error analysis in multibeam hydrographic survey system. South African Journal of Geomatics, 13(2), 236–250. https://dx.doi.org/10.4314/sajg.v13i2.2
Dierssen, H. M., Theberge, A., & Wang, Y. (2014). Bathymetry: History of seafloor mapping. Encyclopedia of Natural Resources, 2, 564. https://doi.org/10.1081/E-ENRW-120047531
Dierssen, H. M., & Theberge, A. E. (2020). Bathymetry: Seafloor mapping history. Coastal and Marine Environments, 195. https://doi.org/10.1201/9780429441004-21
Du, Z., Wells, D. E., & Mayer, L. A. (1996). An approach to automatic detection of outliers in multibeam echo sounding data. The Hydrographic Journal. 79, 19-23. https://doi.org/https://scholars.unh.edu/ccom_affil/11/
Elhassan, I. (2015). Development of bathymetric techniques. Kingdom of Saudi Arabia, FIG Working Week, 1, 2015. https://doi.org/https://www.fig.net/resources/proceedings/fig_proceedings/fig2015/papers/ts04a/TS04A_elhassan_7716.pdf
Feng, Y., Zhang, Z., Zhao, X., Ji, R., & Gao, Y. (2018). Gvcnn: Group-view convolutional neural networks for 3d shape recognition. Proceedings of the IEEE conference on computer vision and pattern recognition, 264-272. http://doi.org/10.1109/CVPR.2018.00035
Ferreira, I. O., de Andrade, L. C., Teixeira, V. G., & Santos, F. C. M. (2022). State of art of bathymetric surveys. Boletim de Ciências Geodésicas, 28(1). https://doi.org/https://doi.org/10.1590/s1982-21702022000100002
Ferreira, I. O., Santos, A. d. P. d., Oliveira, J. C. d., Medeiros, N. d. G., & Emiliano, P. C. (2019). Robust methodology for detection of spikes in multibeam echo sounder data. Boletim de Ciências Geodésicas, 25(3). https://doi.org/https://doi.org/10.1590/s1982-21702019000300014
Gao, R., Xu, T., & Ai, Q. (2017). Research on underwater sound velocity calculation, error correction and positioning algorithms. 2017 Forum on Cooperative Positioning and Service (CPGPS), 39-42. https://doi.org/10.1109/CPGPS.2017.8075094
Graham, B., Engelcke, M., & Van Der Maaten, L. (2018). 3d semantic segmentation with submanifold sparse convolutional networks. In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition , 9224-9232. https://doi.org/10.1109/CVPR.2018.00961
Guenther, G. C. (1982). Improved depth selection in the bathymetric swath survey system (BS3) combined offline processing (COP) program.
He, J. C., Zhang, S. H., Cui, X. D., & Feng, W. (2024). Remote sensing for shallow bathymetry: A systematic review. Earth-Science Reviews, 258. https://doi.org/10.1016/j.earscirev.2024.104957
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep residual learning for image recognition. Proceedings of the IEEE conference on computer vision and pattern recognition, 770-778. https://www.cvfoundation.org/openaccess/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html
Herlihy, D. R., Stepka, T. N., & Rulon, T. D. (1992). Filtering Erroneous Soundings from Multibeam Survey Data. International Hydrographic Review, 69(2), 67–76. https://doi.org/https://journals.lib.unb.ca/index.php/ihr/article/view/26178
Hou, T., Huff, L. C., & Mayer, L. A. (2001). Automatic detection of outliers in multibeam echo sounding data. U.S. Hydrographic Conference 2001, 1-12. https://doi.org/https://scholars.unh.edu/ccom/133/
Hovem, J. M. (2013). Ray trace modeling of underwater sound propagation. In Modeling and measurement methods for acoustic waves and for acoustic microdevices. IntechOpen. https://doi.org/10.5772/55935
Hughes Clarke, J. E. (2017). Multibeam echosounders. Submarine geomorphology, 22-24. https://doi.org/https://link.springer.com/chapter/10.1007/978-3-319-57852-1_3
Hussain, Z., Gimenez, F., Yi, D., & Rubin, D. (2018). Differential data augmentation techniques for medical imaging classification tasks. AMIA annual symposium proceedings, 979-984. https://pmc.ncbi.nlm.nih.gov/articles/PMC5977656/
Irish, J. L., & White, T. E. (1998). Coastal engineering applications of high-resolution lidar bathymetry. Coastal engineering, 35(1-2), 47–71. https://doi.org/Doi10.1016/S0378-3839(98)00022-2
Kalybekova, A. (2025). A review of advancements and applications of Satellite-Derived bathymetry. Engineered Science, 35, 1541. https://doi.org/10.30919/es1541
Ladner, R. W., Elmore, P., Perkins, A. L., Bourgeois, B., & Avera, W. (2017). Automated cleaning and uncertainty attribution of archival bathymetry based on a priori knowledge. Marine Geophysical Research, 38(3), 291–301. https://doi.org/10.1007/s11001-017-9304-9
Le Deunf, J., Debese, N., Schmitt, T., & Billot, R. (2020). A review of data cleaning approaches in a hydrographic framework with a focus on bathymetric multibeam echosounder datasets. Geosciences, 10(7), 254. https://doi.org/https://doi.org/10.3390/geosciences10070254
Li, Z., Peng, Z., Zhang, Z., Chu, Y., Xu, C., Yao, S., García-Fernández, Á. F., Zhu, X., Yue, Y., Levers, A., Zhang, J., & Ma, J. (2023). Exploring modern bathymetry: A comprehensive review of data acquisition devices, model accuracy, and interpolation techniques for enhanced underwater mapping . Frontiers in Marine Science, 10, 1178845. https://doi.org/10.3389/fmars.2023.1178845
Lin, T.-Y., Goyal, P., Girshick, R., He, K., & Dollár, P. (2017). Focal loss for dense object detection. Proceedings of the IEEE international conference on computer vision, 2980-2988. https://openaccess.thecvf.com/content_iccv_2017/html/Lin_Focal_Loss_for_ICCV_2017_paper.html
Lirakis, C., & Bongiovanni, K. (2000). Automated multibeam data cleaning and target detection. OCEANS 2000 MTS/IEEE Conference and Exhibition, 1, 719-723. http://doi.org/10.1109/OCEANS.2000.881336
Lobecker, E., VerPlanck, N., Stuart, E., Peters, C., Forrest, M., Kissinger, K., & Hoy, S. (2010). Mapping data report. EX1001 mapping data report, 10, 01. http://doi.org/10.7289/V5/MDR-OER-EX1001
Long, J. W., Zhang, H. M., & Zhao, J. H. (2023). A Comprehensive Deep Learning-Based Outlier Removal Method for Multibeam Bathymetric Point Cloud. IEEE Transactions on Geoscience and Remote Sensing, 61, 1–22. https://doi.org/Artn 420162210.1109/Tgrs.2023.3242095
Lu, D., Li, H., Wei, Y., & Zhou, T. (2010). Automatic outlier detection in multibeam bathymetric data using robust LTS estimation. 2010 3rd International Congress on Image and Signal Processing, 9, 4032-4036. https://doi.org/10.1109/CISP.2010.5648184
Mann, M., Agathoklis, P., & Antoniou, A. (2001). Automatic outlier detection in multibeam data using median filtering. 2001 IEEE Pacific Rim Conference on Communications, Computers and Signal Processing, 2, 690-693. https://doi.org/10.1109/PACRIM.2001.953726
Mills, J., & Dodd, D. (2014). Ellipsoidally referenced surveying for hydrography. The International Hydrographic Review, 6. https://journals.lib.unb.ca/index.php/ihr/article/view/20887
Mosteller, F., & Tukey, J. W. (1977). Data analysis and regression. A second course in statistics. Addison-Wesley series in behavioral science: quantitative methods.
O’Reilly, C., Parsons, S., & Langelier, D. (1997). A Universal Seamless Vertical Reference Surface for Hydrographic Bathymetry with Specific Application to Seismic Areas. International Association of Geodesy Symposia, 117, 736-743. https://doi.org/10.1007/978-3-662-03482-8_97
Qi, C. R., Su, H., Mo, K., & Guibas, L. J. (2017). Pointnet: Deep learning on point sets for 3d classification and segmentation. Proceedings of the IEEE conference on computer vision and pattern recognition, 652-660. https://doi.org/10.1109/IJCNN.2016.7727386
Qi, C. R., Yi, L., Su, H., & Guibas, L. J. (2017). Pointnet++: Deep hierarchical feature learning on point sets in a metric space. Advances in neural information processing systems, 30. https://dl.acm.org/doi/abs/10.5555/3295222.3295263
Rezvani, M. H., Sabbagh, A., & Ardalan, A. A. (2015). Robust Automatic Reduction of Multibeam Bathymetric Data Based on M-estimators. Marine Geodesy, 38(4), 327–344. https://doi.org/10.1080/01490419.2015.1053639
Schureman, P. (1958). Manual of harmonic analysis and prediction of tides. US Government Printing Office. https://doi.org/https://doi.org/10.5962/bhl.title.38116
Sedaghat, L., Hersey, J., & McGuire, M. P. (2013). Detecting spatio-temporal outliers in crowdsourced bathymetry data. Proceedings of the Second ACM SIGSPATIAL International Workshop on Crowdsourced and Volunteered Geographic Information, 55-62. https://dl.acm.org/doi/abs/10.1145/2534732.2534739
Stephens, D., Smith, A., Redfern, T., Talbot, A., Lessnoff, A., & Dempsey, K. (2020). Using three dimensional convolutional neural networks for denoising echosounder point cloud data. Applied Computing and Geosciences, 5, 100016. https://doi.org/https://doi.org/10.1016/j.acags.2019.100016
Su, H., Maji, S., Kalogerakis, E., & Learned-Miller, E. (2015). Multi-view convolutional neural networks for 3d shape recognition. Proceedings of the IEEE international conference on computer vision, 945-953. https://doi.org/10.48550/arXiv.1505.00880
Thomas, H., Goulette, F., Deschaud, J.-E., Marcotegui, B., & LeGall, Y. (2018). Semantic classification of 3D point clouds with multiscale spherical neighborhoods. 2018 International conference on 3D vision (3DV), 390-398. https://doi.org/10.48550/arXiv.1808.00495
Thomas, H., Qi, C. R., Deschaud, J.-E., Marcotegui, B., Goulette, F., & Guibas, L. J. (2019). Kpconv: Flexible and deformable convolution for point clouds. Proceedings of the IEEE/CVF international conference on computer vision, 6411-6420. https://doi.org/10.1109/ICCV.2019.00651
Uy, M. A., Pham, Q.-H., Hua, B.-S., Nguyen, T., & Yeung, S.-K. (2019). Revisiting point cloud classification: A new benchmark dataset and classification model on real-world data. Proceedings of the IEEE/CVF international conference on computer vision, 1588-1597. https://doi.org/10.48550/arXiv.1908.04616
Wang, S., Zhou, P., Wu, Z., Li, J., & Wei, Y. (2018). Detection and Elimination of Bathymetric Outliers in Multibeam Echosounder System Based on Robust Multi-quadric Method and Median Parameter Model. Journal of Engineering Science & Technology Review, 11(3). https://doi.org/10.25103/jestr.113.10
Wang, Y., Sun, Y., Liu, Z., Sarma, S. E., Bronstein, M. M., & Solomon, J. M. (2019). Dynamic graph cnn for learning on point clouds. ACM Transactions on Graphics (tog), 38(5), 1–12. https://doi.org/10.1145/3326362
Wells, D., Kleusberg, A., & Vanĩcek, P. (2023). A seamless vertical-reference surface for acquisition, management and ECDIS display of hydrographic data. https://doi.org/https://gge.ext.unb.ca/Personnel/Vanicek/Seamless.pdf
Willis, M. J. (1999). Proportional-integral-derivative control. Dept. of Chemical and Process Engineering University of Newcastle, 6, 28. 1887-1893. https://doi.org/10.1109/ICRA.2018.8462926
Wu, B., Wan, A., Yue, X., & Keutzer, K. (2018). Squeezeseg: Convolutional neural nets with recurrent crf for real-time road-object segmentation from 3d lidar point cloud. IEEE International Conference on Robotics and Automation (ICRA), Brisbane, QLD, Australia, 1887-1893. https://doi.org/10.1109/ICRA.2018.8462926
Yan, Y., Mao, Y., & Li, B. (2018). SECOND: Sparsely Embedded Convolutional Detection. Sensors, 18(10), 3337. https://doi.org/10.3390/s18103337
Yang, F., Li, J., Chu, F., & Wu, Z. (2007). Automatic detecting outliers in multibeam sonar based on density of points. OCEANS, 1-4. https://doi.org/10.1109/OCEANSE.2007.4302202
Zhao, H., Jiang, L., Jia, J., Torr, P. H., & Koltun, V. (2021). Point transformer. Proceedings of the IEEE/CVF international conference on computer vision, 16259-16268. https://doi.org/10.1109/ACCESS.2021.3116304
Zhou, Y., & Tuzel, O. (2018). Voxelnet: End-to-end learning for point cloud based 3d object detection. Proceedings of the IEEE conference on computer vision and pattern recognition, 4490-4499. https://doi.org/10.1109/CVPR.2018.00472
Zhu, Q. F., Fan, L., & Weng, N. X. (2024). Advancements in point cloud data augmentation for deep learning: A survey. Pattern recognition, 153, 110532. https://doi.org/10.1016/j.patcog.2024.110532
Zielinski, X. G., Adam. (1999). Precise multibeam acoustic bathymetry. Marine Geodesy, 22(3), 157–167. https://doi.org/10.1080/014904199273434