本研究模擬現實中身分不明的船型目標物入侵我國領海時所量測到含有雜訊的雷達散射截面積(Radar Cross Section, RCS)電磁訊號，並藉由引入降噪技術結合深度學習應用於船型目標物的辨識，期望只需要有整艘船部分角度的RCS數據便能快速進行辨識，因為實際量測環境條件會比模擬情況更嚴苛，本研究討論不同的RCS訓練和測試數據集皆加入不同大小的雜訊，並分別執行兩種降噪方法處理後，分析未降噪與降噪後對神經網路辨識的影響。
本研究先討論如何使用透過CST MIRCROWAVE STUDIO(CST MWS)模擬取得的5種船型目標物的RCS數據集，根據不同的蒐集方法分為角度分集與頻率分集，接著分別對訓練與測試數據集加入不同大小的隨機高斯雜訊來模擬現實量測環境中雜訊干擾的情況，以及透過穩健主成分分析(Robust Principal Component Analysis, RPCA)及Hankel-SVD技術結合獨立成分分析(Independent components analysis, ICA)的兩種降噪技術，並使用卷積神經網路(Convolutional Neural Network，CNN)來進行降噪後的船型目標物RCS數據辨識。
從研究結果顯示，在角度分集RCS數據中，當雜訊係數為10^(-1)和2×10^(-1)以及4×10^(-1)時，應用Hankel-SVD-ICA降噪技術成功辨識率優於RPCA降噪技術，但在應用RPCA降噪技術的成功辨識率在加入不同程度的雜訊係數後皆有穩定的提升;在頻率分集RCS數據中，除了雜訊係數為4×10^(-1)之外，在其他雜訊係數時，應用RPCA降噪技術成功辨識率優於Hankel-SVD-ICA降噪技術。

The purpose of this study is to improve the performance of Convolutional Neural Network(CNN) models for electromagnetic signals recognition of ship targets with noise by using Robust Principal Components Analysis(RPCA) and Hankel Singular Value Decomposition combine Independent Components Analysis(Hankel-SVD-ICA) two noise reduction methods. Radar Cross Section(RCS) of a target is the equivalent area seen by a radar and is also called electromagnetic signature of the object. RCS datasets include several different diversities depending on collection data method. In this study, angle dataset and frequency dataset were adopted. CST MICROWAVE STUDIO(CST MWS), a high-performance 3D electromagnetic analysis software, was used to build RCS datasets by simulating RCS of 5 different ship models. To simulate the realistic measurement with noise, Random number drawn from Gaussian Distributions with different standard deviations were add to the train and test datasets.
The result on angle dataset, the successful recognition rate of applying Hankel-SVD-ICA noise reduction method is better than RPCA noise reduction method when the noise coefficient is 0.1, 0.2 and 0.4. The result on frequency dataset, except for the noise coefficient of 0.4, the successful recognition rate of applying RPCA noise reduction method is better than that of Hankel-SVD-ICA noise reduction method in other noise coefficients.

X. Jiang, A. Hadid, Y. Pang, E. Granger, and X. Feng, Deep Learning in Object Detection and Recognition, Springer Singapore, 2019
I. Harre, RCS in Radar Range Calculations for Maritime Targets, Retrieved on: April. 11, 2022. [Online].
Available: http://www.mar-it.de/Radar/RCS/RCS_xx.pdf
P. Blacksmith, Jr. R. E. Hiatt, and R. B. Mack, “Introduction to radar cross-section measurements,” Proc. IEEE, vol. 53, pp. 901-920, 1965
X. Xu, W. Li, Q. Ran, Q. Du, L. Gao and B. Zhang, "Multisource remote sensing data classification based on convolutional neural network," in IEEE Transactions on Geoscience and Remote Sensing, vol. 56, no. 2, pp. 937-949, Feb 2018.
S. Xing, and S. Zhang, “Ship Model Recognition Based on Convolutional Neural Networks,” IEEE International Conference on Mechatronics and Automation (ICMA), Aug 5-8, Changchun, China, pp. 144-148, 2018.
X. Zhao, X. Xie, X. Quan and X. Sun, "Intelligent recognition method based on deep learning for locally damaged ship targets," 2019 IEEE 8th Data Driven Control and Learning Systems Conference (DDCLS), May 24-27, Dali, Yunnan Province, China, pp. 998-1002, 2019.
E. Julianto, A. Khumaidi, P. Priyonggo, MB. Rahmat, S. Sarena, R. Adhitya, B. Herijono, G. Suharjito and I. Munadhif, "Object recognition on patrol ship using image processing and convolutional neural network (CNN)," International Conference on Applied Science and Technology (ICAST on Engineering Science), Bali, Indonesia, October 24-25 2019.
I. T. Jolliffe, Principal Component Analysis, 2^nded, NY: Springer New York, Inc., 2002.
C. Eckart and G. Young, "The approximation of one matrix by another of lower rank," Psychometrika ,vol. 1,no. 3, pp. 211–218, Sep 1936.
E. J. Candès, J. K. Romberg and T. Tao, "Stable signal recovery from incomplete and inaccurate measurements," Communications on Pure and Applied Mathematics: A Journal Issued by the Courant Institute of Mathematical Sciences, vol. 59, Issue 8, pp. 1207-1223, Mar 2006.
J. Wright, A. Ganesh, S. Rao, Y. Peng and Y Ma, "Robust principal component analysis : exact recovery of corrupted low-rank matrices via convex optimization," Advances in Neural Information Processing Systems 22, 2009.
L. Z. Chen, M. Chen, and Y. Ma. "The augmented lagrange multiplier method for exact recovery of corrupted low-rank matrices," Structural Biology, vol. 181, no. 2, pp. 116-127, Sep 2010.
E. J. Candès, X. Li, Y. Ma, and J. Wright, "Robust principal component analysis ?," Journal of the ACM, vol. 58, Issue 3, no.11, pp. 1-37, May 2011.
Q. Song, G. Yan, G. Tang and F. Ansari, "Robust principal component analysis and support vector machine for detection of microcracks with distributed optical fiber sensors," Mechanical Systems and Signal Processing, vol. 146, no.107019, 2021.
A. Hyvärinen and E. Oja, "Independent component analysis: algorithms and applications," Neural Networks, vol. 13, Issues 4–5, pp. 411-430, 2000.
P. Gao, Ee-C. Chang and L. Wyse, "Blind separation of fetal ECG from single mixture using SVD and ICA," Fourth International Conference on Information, Communications and Signal Processing, 2003 and the Fourth Pacific Rim Conference on Multimedia. Proceedings of the 2003 Joint, vol. 3, pp. 1418-1422.
X. Zhao and B. Ye, "Similarity of signal processing effect between Hankel matrix-based SVD and wavelet transform and its mechanism analysis," Mechanical Systems and Signal Processing, vol. 23, Issue 4, pp. 1062-1075, 2009.
Y. Yue, T. Jiang, C. Han, J. Wang, Y. Chao and Q. Zhou, "Suppression of periodic interference during tunnel seismic predictions via the Hankel-SVD-ICA method," Journal of Applied Geophysics, vol. 168, pp. 107-117, 2019.
A. Krizhevsky, I. Sutskever and G. E. Hinton, "Imagenet classification with deep convolutional neural networks," Communications of the ACM, vol. 60, Issue 6, pp. 84-90, June, 2017.
Y. Grishin and K. Konopko, "Faults diagnosis of stochastic dynamic systems based on neural network probability density function estimation," Proceedings of the 12th IEEE Mediterranean Electrotechnical Conference, Dubrovnik, Croatia, May 12-15, pp. 335-338, 2004.
K. Horiuchi and K. Kameyama, "Parameter density inheritance using kernel density estimation for efficient CNN learning," IEEE International Symposium on Signal Processing and Information Technology (ISSPIT), Louisville, Kentucky, USA, Dec 6-8, pp. 308-313, 2018.