海洋波浪監測技術用於測量與觀測海洋表面之波高、週期及波向等參數，為海象監控與災害預警系統的重要基礎。隨著人工智慧(Artificial Intelligence, AI)領域的發展，深度學習模型已具備處理非線性迴歸(Non-linear Regression)問題的能力，逐漸應用於智慧型海洋監測系統中。其中，光學(Optic)攝影設備因具有建置快速、成本低廉等優勢，將其應用於智慧型波浪監測系統也成為可能。本研究延伸前人所提出之波高辨識模型，進一步建置結合光學海面影像與深度學習技術之波高預測模型，探討其於波高預測的可行性。
本研究採用長期遞迴卷積網路(Long-term Recurrent Convolutional Network, LRCN)融合卷積神經網路(Convolutional Neural Network, CNN)與長短期記憶網路(Long Short Term Memory Network, LSTM)，分別建置波高辨識與預測模型。輸入資料來自海岸光學攝影機拍攝之海面影像，與鄰近浮標量測之波高數據。為強化前人提出之基於海面動態影像的波高辨識模型，本研究將訓練資料擴增，透過特徵正規化與超參數調整以建置波高辨識模型。與前人研究結果相比，本研究優化之辨識模型無因次化均方根誤差由23.4% 降至18.4%，相關係數由0.88提升至0.93，單筆資料辨識速率由1.5秒縮短至0.04秒。顯示本研究優化方法能有效提升辨識準確度與效率。
本研究進一步建置波高預測模型，並針對不同影像時長進行訓練，結果顯示單張影像之驗證誤差最小，表明其已具有足夠之特徵提供模型進行預測。透過優化預測模型超參數與架構以建置最佳化波高預測模型。測試結果顯示，模型預測波高與實測波高間相關係數達0.89，具有良好的預測能力。為評估模型的泛化能力，將其應用於三處不同地點的海岸攝影機影像，無因次化均方根誤差均在三成以內，顯示將本研究模型應用在其他場域也具有預測當地浪況的能力。另外，本研究針對不同預測時間進行分析，隨著預測時間增長誤差逐漸上升，然而無因次均方根誤差仍控制於三成以內，反映長期遞迴卷積網路在波高預測任務中具有穩定的趨勢掌握能力。綜合上述結果，顯示基於光學影像之長期遞迴卷積網路模型於波浪預測具備可行性與發展潛力。

This study integrates AI with optical imaging technology to automatically recognize and predict coastal wave heights. The proposed method adopts a Long-Term Recurrent Convolutional Network (LRCN) architecture, which combines a CNN and a LSTM network to simultaneously capture spatial features from individual frames and temporal dynamics across image sequences.
Compared to previous wave height recognition models, this study enhances performance by expanding the training dataset, applying feature normalization, and tuning hyperparameters. As a result, the model reduces the normalized RMSE from 23.4% to 18.4%, and improves the correlation with observed wave heights from 0.88 to 0.93. In addition, the optimized model significantly improves inference speed, reducing per-frame processing time and increasing computational efficiency, making it suitable for potential real-time applications.
The study also develops a prediction model using single-frame inputs, which achieves a correlation coefficient of 0.89, demonstrating strong predictive capability. When tested on coastal images from different locations, the model shows good generalization and adaptability to various environments. Overall, the LRCN-based approach proposed in this study provides a feasible and effective solution for coastal wave height estimation and forecasting, offering high accuracy and computational efficiency.

王敘民、董東璟、蔡政翰、林芳如(2022)，海岸裂流觀測之研究，中國土木水利工程學刊，第四十九卷，第六期，第24-30頁。
交通部中央氣象署(2019)，異常海象機率預警研究與作業試用，研究計畫報告。
陳威成、楊文榮、傅仲偉、董東璟、廖建明(2022)，人工智慧技術在海洋觀測之應用，中國土木水利工程學刊，第四十九卷，第六期，第31-39頁。
陳盈智、董東璟、林芳如、滕春慈(2022)，海岸異常波浪觀測之研 究，中國土木水利工程學刊，第四十九卷，第六期，第40-45頁。
許家瑄(2023)，近五十年台灣周遭海域與北南海波候變化分析。國立成功大學碩士論文。
楊大成、范揚洺、李汴軍、高家俊、莊士賢、滕春慈(2010)，氣候變遷對臺灣東北角海域波候變化之初步研究，第32屆海洋工程研討會論文集，第155-160頁。
葉浩君(2020)，臺灣鄰近海域湧浪研究。國立成功大學碩士論文。
劉彧愷(2024)，以長期遞迴卷積網路(LRCN)建置即時波高辨識模型。國立成功大學碩士論文。
Amari, S. (1993). Backpropagation and stochastic gradient descent method. Neurocomputing, 5(4–5), 185–196.
Ardhuin, F., Stopa, J. E., Chapron, B., Collard, F., Husson, R., Jensen, R. E., Johannessen, J., Mouche, A., Passaro, M., Quartly, G. D., Swail, V., & Young, I. (2019). Observing sea states. Frontiers in Marine Science, 6.
Ba, J. L., Kiros, J. R., & Hinton, G. E. (2016). Layer normalization. arXiv preprint arXiv:1607.06450.
Berbić, J., Ocvirk, E., Carević, D., & Lončar, G. (2017). Application of neural networks and support vector machine for significant wave height prediction. Oceanologia, 59(3), 331–349.
Booij, N., Ris, R. C., Holthuijsen, L. H., & Faculty of Civil Engineering, Delft University of Technology, Delft, Netherlands. (1999). A third-generation wave model for coastal regions, Part I, Model description and validation. Journal of Geophysical Research Atmospheres.
Cavaleri, L., Alves, J., Ardhuin, F., Babanin, A., Banner, M., Belibassakis, K., Benoit, M., Donelan, M., Groeneweg, J., Herbers, T., Hwang, P., Janssen, P., Janssen, T., Lavrenov, I., Magne, R., Monbaliu, J., Onorato, M., Polnikov, V., Resio, D., . . . Young, I. (2007). Wave modelling – The state of the art. Progress in Oceanography, 75(4), 603–674.
Chang, Z., Zhang, Y., & Chen, W. (2018). Effective Adam-Optimized LSTM Neural Network for Electricity Price Forecasting. 2018 IEEE 9th International Conference on Software Engineering and Service Science (ICSESS), 245–248.
Choi, H., Park, M., Son, G., Jeong, J., Park, J., Mo, K., & Kang, P. (2020). Real-time significant wave height estimation from raw ocean images based on 2D and 3D deep neural networks. Ocean Engineering, 201, 107129.
Cooijmans, T., Ballas, N., Laurent, C., Gülçehre, Ç., & Courville, A. (2016). Recurrent batch normalization. arXiv preprint arXiv:1603.09025.
Demetriou, D., Michailides, C., Papanastasiou, G., & Onoufriou, T. (2021). Coastal zone significant wave height prediction by supervised machine learning classification algorithms. Ocean Engineering, 221, 108592.
Donahue, J., Hendricks, L. A., Rohrbach, M., Venugopalan, S., Guadarrama, S., Saenko, K., & Darrell, T. (2016). Long-Term recurrent convolutional networks for visual recognition and description. IEEE Transactions on Pattern Analysis and Machine Intelligence, 39(4), 677–691.
Ellenson, A. N., Simmons, J. A., Wilson, G. W., Hesser, T. J., & Splinter, K. D. (2020). Beach state recognition using ARGUs imagery and convolutional neural networks. Remote Sensing, 12(23), 3953.
Graves, A., Mohamed, A., & Hinton, G. (2013). Speech recognition with deep recurrent neural networks. IEEE International Conference on Acoustics Speech and Signal Processing, 6645–6649.
Greff, K., Srivastava, R. K., Koutnik, J., Steunebrink, B. R., & Schmidhuber, J. (2016). LSTM: a Search space Odyssey. IEEE Transactions on Neural Networks and Learning Systems, 28(10), 2222–2232.
He, K., Zhang, X., Ren, S., & Sun, J. (2016). Deep Residual Learning for Image Recognition. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 770–778.
Hinton, G., Nitish Srivastava, & Kevin Swersky. (2012). Neural networks for machine learning.
Hochreiter, S., & Schmidhuber, J. (1997). Long Short-Term memory. Neural Computation, 9(8), 1735–1780.
Huang, Q., & Cui, Z. (2023). Study on prediction of ocean effective wave height based on hybrid artificial intelligence model. Ocean Engineering, 289, 116137.
Huber, P. J. (1964). Robust estimation of a location parameter. The Annals of Mathematical Statistics, 35(1), 73–101.
Huang, Y. (2020). Research status and applications of nature-inspired algorithms for agri-food production. International Journal of Agricultural and Biological Engineering, 13(4), 1–9.
Ioffe, S., & Szegedy, C. (2015). Batch normalization: Accelerating deep network training by reducing internal covariate shift. arXiv preprint arXiv:1502.03167.
Kang, B., & Vinent, O. D. (2023). The application of CNN-Based image segmentation for tracking coastal erosion and Post-Storm recovery. Remote Sensing, 15(14), 3485.
Kim, Y., Cho, S., & Lee, P. (2023). Wave height classification via deep learning using monoscopic ocean videos. Ocean Engineering, 288, 116002.
Kingma, D. P., & Ba, J. L. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.
Kwon, J., Chang, W., & Yang, Y. J. (2023). Significant wave height prediction from X-band marine radar images using deep learning with 3D convolutions. PLoS ONE, 18(10), e0292884.
Loizou, A., & Christmas, J. (2021). Sea State Estimation from Uncalibrated, Monoscopic Video. SN Computer Science, 2(4).
Loshchilov, I., & Hutter, F. (2017). Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101.
Mao, M., Lee, A., & Hong, M. (2024). Deep Learning Innovations in Video Classification: A survey on techniques and dataset evaluations. Electronics, 13(14), 2732.
Massa, L., Barbosa, A., Oliveira, K., & Vieira, T. (2020). LRCN-RetailNet: A recurrent neural network architecture for accurate people counting. Multimedia Tools and Applications, 80(4), 5517–5537.
Okewu, E., Adewole, P., & Sennaike, O. (2019). Experimental comparison of stochastic optimizers in deep learning. In Lecture notes in computer science (pp. 704–715).
Sen, S. Y., & Ozkurt, N. (2020). Convolutional Neural Network Hyperparameter Tuning with Adam Optimizer for ECG Classification. 2022 Innovations in Intelligent Systems and Applications Conference (ASYU), 1–6.
Shukla, R. P., Kinter, J. L., & Shin, C. (2018). Sub-seasonal prediction of significant wave heights over the Western Pacific and Indian Oceans, part II: The impact of ENSO and MJO. Ocean Modelling, 123, 1–15.
Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for Large-Scale image recognition. arXiv preprint arXiv:1409.1556.
Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdinov, R., & Department of Computer Science, University of Toronto. (2014). Dropout: A Simple Way to Prevent Neural Networks from Overfitting. In Y. Bengio (Ed.), Journal of Machine Learning Research (Vol. 15, pp. 1929–1958).
Tibshirani, R. (1996). Regression shrinkage and selection via the lasso. Journal of the Royal Statistical Society Series B (Statistical Methodology), 58(1), 267–288.
WAMDI Group. (1988). The WAM Model—A third generation ocean wave prediction model. Journal of Physical Oceanography, 18(12), 1775–1810.
Wang, L., Deng, X., Ge, P., Dong, C., Bethel, B. J., Yang, L., & Xia, J. (2022). CNN-BiLSTM-Attention Model in Forecasting Wave Height over South-East China Seas. Computers, Materials & Continua/Computers, Materials & Continua (Print), 73(1), 2151–2168.
Wang, Y., He, Z., & Wang, L. (2021). Truck driver fatigue detection based on video sequences in Open-Pit mines. Mathematics, 9(22), 2908.
WAVEWATCH III Development Group. (2019). User Manual and System Documentation of WAVEWATCH III Version 6.07 (Vol. 333).
Wei, L., Guan, L., Qu, L., & Guo, D. (2020). Prediction of sea surface temperature in the China seas based on Long Short-Term Memory Neural Networks. Remote Sensing, 12(17), 2697.
Xiong, R., Yang, Y., He, D., Zheng, K., Zheng, S., Xing, C., Zhang, H., Lan, Y., Wang, L., & Liu, T. (2020). On layer normalization in the transformer architecture. arXiv preprint arXiv:2002.04745.
Zhang, C., & Li, M. (2024). Three-stage ocean wave elements extraction using deep learning based on in-situ monocular videos from offshore infrastructure. Ocean Engineering, 313, 119356.
Zhang, X., Li, Y., Gao, S., & Ren, P. (2021). Ocean Wave Height Series Prediction with Numerical Long Short-Term Memory. Journal of Marine Science and Engineering, 9(5), 514.