我們常利用波動來對未知的目標物或者難以到達的區域做探測，而電磁波就是其中的一種常用的波動，能夠透過電磁遙測推估目標物的訊息，例如敵方的軍艦或軍機，旅客行李箱亦或者建築物水泥中有多少鋼筋、探測泥土的含水量等。對於這些難以肉眼辨識且直接觸碰的目標，使用電磁波來探測是很合適的方法，但要做到電磁探測不是一件容易的事情，必須先利用馬克斯威方程式與正散射理論計算出在入射光條件已知下，什麼特性的目標物例如形狀、材料，會有怎樣的散射反應，散射是任意物體皆會在與電磁波接觸時會有的電磁現象，當我們能夠掌握散射場與目標物之關係，就能皆從收集散射場推估目標物特性，以往都是以演算法來實現，例如螢火蟲演算法或者粒子群演算法等，而本研究是深度學習的方式做逆散射，且選定卷積神經網路，以介質柱目標截面形狀為特性，找出散射場與介質柱目標截面之關係。
在第二章先介紹介質目標物的正散射理論，如何在入射電場各項條件已知的狀況下算出2D介質柱截面之散射場，並探討各項在電算機實現正散射理論之方法，第三章先利用卷積神經網路做散射場迴歸模擬，再使用卷積神經網路做逆散射實現，也就是透過大量的目標物形狀與其對應的散射反應做卷積神經模型訓練，如此實現目標物形狀迴歸之應用。並利用2種不同的資料排列方式比較訓練效用之差別，且使用適當的模型評估指標各自呈現結果。

We often use waves to detect unknown targets or hard-to-reach areas, and electromagnetic waves are one of waves we usually to use. We are able to estimate target information through electromagnetic remote sensing. For these targets which are difficult to be recognized by the naked eye, the way of using electromagnetic waves to detect is a very suitable method. Scattering is an electromagnetic phenomenon that any object will have when it comes into contact with electromagnetic waves. Once we get the relationship between the scattered field and the target, we will get the characteristics of the target from the collected scattered field. In the past, it was implemented by algorithms, such as firefly algorithm or particle swarm algorithm, etc. In this study, inverse scattering is performed in a deep learning method and its target object is characterized by the cross-sectional shape of the dielectric cylinder to find the relationship between the scattering field and the cross-section of the dielectric cylinder.we use the convolutional neural network for the scattering field regression and realize inverse scattering with the convolutional neural network. In other words, we train model through a large number of target shapes and their corresponding scattering reactions to get the application of scattering field and target shape regression.Finally, we compare result in two different data arrangement methods and use appropriate model index to evaluate the performance.

[1] M. Peña-Cabrera, V. Lomas, and G. Lefranc, "Fourth industrial revolution and its impact on society," in 2019 IEEE CHILEAN Conference on Electrical, Electronics Engineering, Information and Communication Technologies (CHILECON), pp. 1-6 Nov. 2019.
[2] J. M. Tien, "Towards the next industrial revolution," in 2012 4th IEEE International Symposium on Logistics and Industrial Informatics, pp. 13-14,Sept. 2012.
[3] T. Guarda et al., "Internet of Things challenges," in 2017 12th Iberian Conference on Information Systems and Technologies (CISTI), pp. 1-4, June 2017.
[4] A. I. Galushkin, "Continual neural networks," in Proceedings of 1993 International Conference on Neural Networks (IJCNN-93-Nagoya, Japan) , pp. 395-398 vol.1, Oct. 1993.
[5] J. M. Bishop and R. J. Mitchell, "Neural networks-an introduction," in IEE Colloquium on Neural Networks for Systems: Principles and Applications1, pp. 1/1-1/3, Jan. 1991.
[6] M. K. James, L. Derong, and B. F. David, "Introduction and Single-Layer Neural Networks," in Fundamentals of Computational Intelligence: Neural Networks, Fuzzy Systems, and Evolutionary Computation: IEEE, pp. 5-34,2016
[7] Z. Nazarchuk, R. Hryniv, M. Shahin, and A. Synyavskyy, "A method of inverse scattering problem solution for penetrable objects using back-scattering data," in 2015 XXth IEEE International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), pp. 23-26, Sept. 2015.
[8] W. Ding and W. Fang, "Target Tracking by Sequential Random Draft Particle Swarm Optimization Algorithm," in 2018 IEEE International Smart Cities Conference (ISC2), pp. 1-7,Sept. 2018.
[9] P. Guo, X. Wang, and Y. Han, "The enhanced genetic algorithms for the optimization design," in 2010 3rd International Conference on Biomedical Engineering and Informatics, vol. 7, pp. 2990-2994, Oct. 2010.
[10] T. B. A. Senior, "Analytical techniques in scattering," in IEEE Antennas and Propagation Society International Symposium. Transmitting Waves of Progress to the Next Millennium. 2000 Digest. Held in conjunction with: USNC/URSI National Radio Science Meeting (C , vol. 3, p. 1637 July 2000.
[11] H. Karl, "Scattering Theory," in Advanced Theory of Semiconductor Devices: IEEE, pp. 89-108, 2000.
[12] B.-L. Olga, M. L. Victor, D. D. Amy, P. Byung-Kwon, and S. Aditya, "Radar Principles," in Doppler Radar Physiological Sensing: IEEE, pp. 21-38,2016.
[13] R. Curiac and I. R. Ciric, "Analysis of wave scattering by a lossy dielectric using single source surface integral equations," in IEEE CCECE2002. Canadian Conference on Electrical and Computer Engineering. Conference Proceedings (Cat. No.02CH37373), vol. 1, pp. 297-300 ,May 2002.
[14] H. P. Christos, "Mom," in Turing (A Novel about Computation): MITP, pp. 53-57, 2005.
[15] P. Raffaele, "The 2D Scattering Equations for Dielectric Targets," in Introduction to Ground Penetrating Radar: Inverse Scattering and Data Processing: IEEE, pp. 48-78., 2014.
[16] F. Q. Lauzon, "An introduction to deep learning," in 2012 11th International Conference on Information Science, Signal Processing and their Applications (ISSPA) , pp. 1438-1439, July 2012.
[17] F. Ertam and G. Aydın, "Data classification with deep learning using Tensorflow," in 2017 International Conference on Computer Science and Engineering (UBMK), pp. 755-758, Oct. 2017.
[18] Y. Zhang, W. Ding, and C. Liu, "Summary of Convolutional Neural Network Compression Technology," in 2019 IEEE International Conference on Unmanned Systems (ICUS), pp. 480-483, Oct. 2019.
[19] I. Akira, "Inverse Scattering," in Electromagnetic Wave Propagation, Radiation, and Scattering: From Fundamentals to Applications: IEEE, pp. 553-585,2017.
[20] K.Barkeshli,AdvancedElectromagnetics and Scattering Theory ,Springer,2015,ch3.
[21] K.Barkeshli,AdvancedElectromagnetics and Scattering Theory ,Springer,2015,ch8.
[22] K.Barkeshli,AdvancedElectromagnetics and Scattering Theory ,Springer,2015,ch9.
[23] H.Zhou and W.Hong,Improvement of MoM-CG-FFT scheme for EM scattering From an infinite dielectric cylinder,Journal of Southeast:Natural Science, 2002.
[24] A. Faize,A. Driouach,A. Kaabal,G. Alsharahi and A. M. Qasem ,MOM Application for Calculating the RCS Dielectrics and Arbitrary Two-Dimensional Geometric Shape Formulation of Integral Equations Cylindrical Dielectric,IJEAT,2013.
[25] R. Curiac and I. R. Ciric, "Analysis of wave scattering by a lossy dielectric using single source surface integral equations," in IEEE CCECE2002. Canadian Conference on Electrical and Computer Engineering. Conference Proceedings (Cat. No.02CH37373) , vol. 1, pp. 297-300,May 2002.