A Convolutional Neural Network with Multifrequency and Structural Similarity Loss Functions for Electromagnetic Imaging

• Published in: Sensors (Basel, Switzerland) Vol. 24; no. 15; p. 4994
• Main Authors: Chiu, Chien-Ching, Lin, Che-Yu, Chi, Yu-Jen, Hsu, Hsiu-Hui, Chen, Po-Hsiang, Jiang, Hao
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 01.08.2024
• Subjects: Accuracy; Algorithms; anisotropic objects; Anisotropy; Artificial intelligence; Back propagation; back-propagation scheme; Computational linguistics; convolutional neural network; Deep learning; Electric properties; electromagnetic imaging; Electromagnetism; Fourier transforms; Iterative methods; Language processing; loss function; Natural language interfaces; Neural networks; Optimization; Wavelet transforms; loss function; back-propagation scheme; anisotropic objects; electromagnetic imaging; convolutional neural network; artificial intelligence
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s24154994

In this paper, artificial intelligence (AI) technology is applied to the electromagnetic imaging of anisotropic objects. Advances in magnetic anomaly sensing systems and electromagnetic imaging use electromagnetic principles to detect and characterize subsurface or hidden objects. We use measured multifrequency scattered fields to calculate the initial dielectric constant distribution of anisotropic objects through the backpropagation scheme (BPS). Later, the estimated multifrequency permittivity distribution is input to a convolutional neural network (CNN) for the adaptive moment estimation (ADAM) method to reconstruct a more accurate image. In the meantime, we also improve the definition of loss function in the CNN. Numerical results show that the improved loss function unifying the structural similarity index measure (SSIM) and root mean square error (RMSE) can effectively enhance image quality. In our simulation environment, noise interference is considered for both TE (transverse electric) and TM (transverse magnetic) waves to reconstruct anisotropic scatterers. Lastly, we conclude that multifrequency reconstructions are more stable and precise than single-frequency reconstructions.