ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering

Emre Işcan; Vasil Tabatadze; Kamil Karaçuha; Ertuǧrul Karaçuha

doi:10.1109/DIPED66951.2025.11194410

ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering

Emre Işcan^*
, Vasil Tabatadze
, Kamil Karaçuha
, Ertuǧrul Karaçuha

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

This paper proposes an artificial neural network (ANN)-based method for quickly determining the orientation of a dielectric ellipsoid in 3D electromagnetic pulse scattering. The Method of Auxiliary Sources (MAS) is used to perform the scattering analysis efficiently by handling field singularities and reducing computation time. Gaussian radio pulse serves as the excitation, and the scattered signal is measured at a single observation point. A dataset covering a wide range of orientation angles is used to train the ANN, which accurately estimates the ellipsoid's orientation with an average error of less than 1° for both angles using just one measurement. This approach is more efficient and practical than conventional radar techniques, especially for real-time applications like target tracking and missile detection. This work presents a significant advance by integrating machine learning with electromagnetic analysis to speed up 3D object recognition.

Original language	English
Title of host publication	2025 IEEE 30th International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Proceedings
Publisher	IEEE Computer Society
Pages	117-122
Number of pages	6
ISBN (Electronic)	9798331588144
DOIs	https://doi.org/10.1109/DIPED66951.2025.11194410
Publication status	Published - 2025
Event	30th IEEE International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Tbilisi, Georgia Duration: 8 Sept 2025 → 10 Sept 2025

Publication series

Name	Proceedings of International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED
ISSN (Print)	2165-3585
ISSN (Electronic)	2165-3593

Conference

Conference	30th IEEE International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025
Country/Territory	Georgia
City	Tbilisi
Period	8/09/25 → 10/09/25

Bibliographical note

Publisher Copyright:
© 2025 IEEE.

Keywords

3D electromagnetic diffraction
artificial neural networks
Gaussian pulse feedback
MAS
orientation estimation
real-time identification

Access to Document

10.1109/DIPED66951.2025.11194410

Cite this

Işcan, E., Tabatadze, V., Karaçuha, K., & Karaçuha, E. (2025). ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering. In 2025 IEEE 30th International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Proceedings (pp. 117-122). (Proceedings of International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED). IEEE Computer Society. https://doi.org/10.1109/DIPED66951.2025.11194410

Işcan, Emre ; Tabatadze, Vasil ; Karaçuha, Kamil et al. / ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering. 2025 IEEE 30th International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Proceedings. IEEE Computer Society, 2025. pp. 117-122 (Proceedings of International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED).

@inproceedings{34316a5a3446497b94c6fa7a793f0974,

title = "ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering",

abstract = "This paper proposes an artificial neural network (ANN)-based method for quickly determining the orientation of a dielectric ellipsoid in 3D electromagnetic pulse scattering. The Method of Auxiliary Sources (MAS) is used to perform the scattering analysis efficiently by handling field singularities and reducing computation time. Gaussian radio pulse serves as the excitation, and the scattered signal is measured at a single observation point. A dataset covering a wide range of orientation angles is used to train the ANN, which accurately estimates the ellipsoid's orientation with an average error of less than 1° for both angles using just one measurement. This approach is more efficient and practical than conventional radar techniques, especially for real-time applications like target tracking and missile detection. This work presents a significant advance by integrating machine learning with electromagnetic analysis to speed up 3D object recognition.",

keywords = "3D electromagnetic diffraction, artificial neural networks, Gaussian pulse feedback, MAS, orientation estimation, real-time identification",

author = "Emre I{\c s}can and Vasil Tabatadze and Kamil Kara{\c c}uha and Ertuǧrul Kara{\c c}uha",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 30th IEEE International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 ; Conference date: 08-09-2025 Through 10-09-2025",

year = "2025",

doi = "10.1109/DIPED66951.2025.11194410",

language = "English",

series = "Proceedings of International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED",

publisher = "IEEE Computer Society",

pages = "117--122",

booktitle = "2025 IEEE 30th International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Proceedings",

address = "United States",

}

Işcan, E , Tabatadze, V , Karaçuha, K & Karaçuha, E 2025, ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering. in 2025 IEEE 30th International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Proceedings. Proceedings of International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED, IEEE Computer Society, pp. 117-122, 30th IEEE International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025, Tbilisi, Georgia, 8/09/25. https://doi.org/10.1109/DIPED66951.2025.11194410

ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering. / Işcan, Emre ; Tabatadze, Vasil ; Karaçuha, Kamil et al.
2025 IEEE 30th International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Proceedings. IEEE Computer Society, 2025. p. 117-122 (Proceedings of International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering

AU - Işcan, Emre

AU - Tabatadze, Vasil

AU - Karaçuha, Kamil

AU - Karaçuha, Ertuǧrul

PY - 2025

Y1 - 2025

N2 - This paper proposes an artificial neural network (ANN)-based method for quickly determining the orientation of a dielectric ellipsoid in 3D electromagnetic pulse scattering. The Method of Auxiliary Sources (MAS) is used to perform the scattering analysis efficiently by handling field singularities and reducing computation time. Gaussian radio pulse serves as the excitation, and the scattered signal is measured at a single observation point. A dataset covering a wide range of orientation angles is used to train the ANN, which accurately estimates the ellipsoid's orientation with an average error of less than 1° for both angles using just one measurement. This approach is more efficient and practical than conventional radar techniques, especially for real-time applications like target tracking and missile detection. This work presents a significant advance by integrating machine learning with electromagnetic analysis to speed up 3D object recognition.

AB - This paper proposes an artificial neural network (ANN)-based method for quickly determining the orientation of a dielectric ellipsoid in 3D electromagnetic pulse scattering. The Method of Auxiliary Sources (MAS) is used to perform the scattering analysis efficiently by handling field singularities and reducing computation time. Gaussian radio pulse serves as the excitation, and the scattered signal is measured at a single observation point. A dataset covering a wide range of orientation angles is used to train the ANN, which accurately estimates the ellipsoid's orientation with an average error of less than 1° for both angles using just one measurement. This approach is more efficient and practical than conventional radar techniques, especially for real-time applications like target tracking and missile detection. This work presents a significant advance by integrating machine learning with electromagnetic analysis to speed up 3D object recognition.

KW - 3D electromagnetic diffraction

KW - artificial neural networks

KW - Gaussian pulse feedback

KW - MAS

KW - orientation estimation

KW - real-time identification

UR - https://www.scopus.com/pages/publications/105020786540

U2 - 10.1109/DIPED66951.2025.11194410

DO - 10.1109/DIPED66951.2025.11194410

M3 - Conference contribution

AN - SCOPUS:105020786540

T3 - Proceedings of International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED

SP - 117

EP - 122

BT - 2025 IEEE 30th International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Proceedings

PB - IEEE Computer Society

T2 - 30th IEEE International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025

Y2 - 8 September 2025 through 10 September 2025

ER -

Işcan E , Tabatadze V , Karaçuha K , Karaçuha E. ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering. In 2025 IEEE 30th International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED 2025 - Proceedings. IEEE Computer Society. 2025. p. 117-122. (Proceedings of International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory, DIPED). doi: 10.1109/DIPED66951.2025.11194410

ANN-Based Method for Real-Time Recognition of Dielectric Ellipsoid Orientation in 3D Electromagnetic Pulse Scattering

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this