Influence of denoising of input signal on classification of object location by artificial neural network in ultrawideband radiointroscopy
Abstract
Background: A wide usage of impulse ultrawideband subsurface radars for a number of practical approaches in archeology, construction and humanitarian demining is holding back because of presence of noises and clutters of high level in the reflected field. It often makes the object classification practically unreal for at not big depths and distances from receiving and transmitting antennas. Besides of using special antenna system designs to improve recognition results, it is interesting to apply modern digital signal filtering techniques.
Objectives: To investigate the influence of denoising on the quality of artificial neural network recognition of subsurface objects and their coordinates for a model of additive gaussian noise of a different noise level.
Materials and methods: In this paper the idea of improving the stability of recognition of hidden objects in the presence of outside noise by previous processing of input signals with the latest popular noise reduction methods, such as the caterpillar method and wavelet transform method is verified. To eliminate the randomness of the result of the neural network response for each realization of the additive noise of a given level, a sufficient number of attempts are calculated for each of the methods, and statistics are provided to illustrate the effectiveness of each of the approaches. To check the hypothesis of the efficiency of input signal denoising the numerical simulation of the model of a real ground surface with subsurface object is carried out by means of Finite Difference Time Domain method (FDTD). The artificial neural network is trained on the obtained ideal time dependences of the amplitudes of the reflected field to correctly recognize the position of the object. The training is subsequently checked on the same input signals with additional noise of a certain level. Recognition errors in the last case are compared with similar errors when popular noise reduction procedures are applied to noisy input signals.
Results: It is demonstrated that artificial neural networks have good approximating properties capable to effectively resist the noises in the input signals It is shown that for all noise levels, the caterpillar method statistically degrades the quality of an object recognition. The wavelet-transform method statistically improves slightly the classification of objects than for absence of denoising, but this result is not stable.
Conclusion: For effective application of methods of noise filtration in received signals of impulse radar it is nessusary to have previous knowledge about noise character or peculiarities of useful signal. Implementation of denoising techniques without the use of this knowledge cannot improve the recognition quality of surface objects.
Downloads
References
Daniels DJ. Ground penetrating radar, 2nd ed. London; IEEE; 2004. 726 p.
Harmuth H. Nonsinusoidal waves for radar and radiocommunications. New York: Academic Press; 1981. 396 p.
Taylor JD. Ultrawidebandradar: applications and design. Boca Raton, NewYork: CRC Press; 2012. 536 p.
Ristic A, Govedarica M, Vrtunski M, Pctrovacki D. Application of GPR for creating underground structure model of specific areas of interest. Proceedings of the 15th International Conference on Ground Penetrating Radar; 2014 Jun 30- Jul 4; Brussels; p. 450-455.
Liu H, Huang X, Xing B, Cui J, Spencer BF, Liu QH. Estimating Azimuth of Subsurface Linear Targets By Polarimetric GPR. Proceedings of the IEEE International Geoscience and Remote Sensing Symposium; 2018 Jul 22-27; Valencia; p. 6784-6787.
Tivive FHC, Bouzerdoum A, Abeynayake C. GPR Target Detection by Joint Sparse and Low-Rank Matrix Decomposition. IEEE Transactions on Geoscience and Remote Sensing. 2018 of November;57(5):2583-2595.
Zhao S, Al-Qadi IL. Super-Resolution of 3-D GPR Signals to Estimate Thin Asphalt Overlay Thickness Using the XCMP Method. IEEE Transactions on Geoscience and Remote Sensing. 2018 Aug;57(2):893-901.
Taflove A, Hagness S. Computational Electrodynamics: The Finite-Difference Time-Domain Method, 3rd ed. Boston, London; Artech House; 2005. Chapter 9, Computational Electromagnetics: The Finite-Difference Time-Domain Method; p. 629-669.
Liu Y, Guo LX. FDTD investigation on GPR detecting of underground subsurface layers and buried objects. Proceedings of the IEEE MTT-S International Conference on Numerical Electromagnetic and Multiphysics Modeling and Optimization; 2016 Jul 27-29; Beijing; p. 1-2. doi: 10.1109/NEMO.2016.7561622
Giannakis I, Giannopoulos A, Warren C. A Machine Learning-Based Fast-Forward Solver for Ground Penetrating Radar With Application to Full-Waveform Inversion. IEEE Transactions on Geoscience and Remote Sensing. 2019 January;57(7):4417-4426.
Hasija K, Vadhera S, Kumar A, Kishore A. Detection and location of faults in underground cable using Matlab/Simulink/ANN and OrCad. Proceedings of the 6th IEEE Power India International Conference; 2014 Dec 5-7; Delhi. p. 1-5. doi: 10.1109/POWERI.2014.7117624
Birkenfeld S. Automatic detection of reflexion hyperbolas in gpr data with neural networks. Proceedings of the World Automation Congress; 2010 Sep 19-23; Kobe;. p. 1-6.
Shyrokorad D, Dumin O, Dumina O, Katrich V, Chebotarev V. Approximating properties of artificial neural network in time domain for the analysis of electromagnetic fields reflected from model of human body surface. Proceedings of the Microwaves, Millimeter and Submillimeter Waves; 2010 Jun 21-26; Kharkiv; p. 1-3. doi: 10.1109/MSMW.2010.5546075
Moon WM, Miao XG, Singhroy V, Lowman PD. Integrated geophysical imaging of the Sudbury Structure using ERS-1 SAR and other airborne geophysical data. Proceedings of the IEEE International Geoscience and Remote Sensing Symposium; 1994 Aug 8-12; Pasadena; p. 711.
Zhou Y, Chen W. MCA-Based Clutter Reduction From Migrated GPR Data of Shallowly Buried Point Target. IEEE Transactions on Geoscience and Remote Sensing. 2008 September;57(1):432-448.
Bai X, Peng X. Radar Image Series Denoising of Space Targets Based on Gaussian Process Regression. IEEE Transactions on Geoscience and Remote Sensing. 2009 Feb;57(7):4659-4669.
Dumin O, Prishchenko O, Pochanin G, Plakhtii V, Shyrokorad D. Subsurface Object Identification by Artificial Neural Networks and Impulse Radiolocation. Proceedings of the IEEE Second International Conference on Data Stream Mining; 2018 Aug 21-25; Lviv; p. 434-437.
Ogurtsova T, Ruban V, Pojedinchuk A, Pochanin O, Pochanin G, Capineri L, Falorni P, Borgioli G, Bechtel T, Crawford F. Criteria for Selecting Object Coordinates at Probing by the Impulse UWB GPR with the “1Tx + 4Rx” Antenna System. Proceedings of the International Conference on Ultrawideband and Ultrashort Impulse Signals; 2018 Sep 4-7; Odessa; p. 161-164.
Varianytsia-Roshchupkina LA, Roshchupkin SV. Subsurface object imaging with two types of RTR-differential GPR system. Proceedings of the 8th International Conference on Ultrawideband and Ultrashort Impulse Signals; 2016 Sep 5-11; Odessa; p. 145-147.
Haykin S. Neural Networks, 2nd ed, New Jersey: Prentice-Hall; 1999. p. 823.
Shyrokorad D, Dumin O, Dumina O. Time domain analysis of reflected impulse fields by artificial neural network. Proceedings of the 4th International Conference on Ultrawideband and Ultrashort Impulse Signals; 2008 Sep 15-19; Sevastopol; p. 124-126.
Dumin O, Dumina O, Shyrokorad D. Time domain analysis of fields reflected from model of human body surface using artificial neural network. Proceedings of the 3rd European Conference on Antennas and Propagation; 2009 Mar 23-27; Berlin; p. 235-238.
Shyrokorad D, Dumin O, Dumina O, Katrich V. Analysis of transient fields reflected from model of human body surface using convolutional neural network. Proceedings of the International Conference on Mathematical Methods in Electromagnetic Theory; 2010 Sep 6-8; Kyiv; p. 1-4. doi: 10.1109/MMET.2010.5611389.
Dumin OM, Prishchenko O, Shyrokorad D, Plakhtii V. Application of UWB Electromagnetic Waves for Subsurface Object Location Classification by Artificial Neural Networks. Proceedings of the 9th International Conference on Ultrawideband and Ultrashort Impulse Signals; 2018 Sep 4-7; Odessa; p. 290-293.
