Microwave Imaging by Deep Learning Network: Feasibility and Training Method

Research output: Contribution to journalArticlepeer-review

Abstract

Microwave image reconstruction based on a deep learning method is investigated in this article. The neural network is capable of converting measured microwave signals acquired from a 24 \times 24 antenna array at 4 GHz into a 128 \times 128 image. To reduce the training difficulty, we first developed an autoencoder by which high-resolution images ( 128 \times 128 ) were represented with 256 \times 1 vectors; then we developed the second neural network which aimed to map microwave signals to the compressed features ( 256 \times 1 vector). Two neural networks can be combined to a full network to make reconstructions, when both are successfully developed. The present two-stage training method reduces the difficulty in training deep learning networks (DLNs) for inverse reconstruction. The developed neural network is validated by simulation examples and experimental data with objects in different shapes/sizes, placed in different locations, and with dielectric constant ranging from 2 to 6. Comparisons between the imaging results achieved by the present method and two conventional approaches: distorted Born iterative method (DBIM) and phase confocal method (PCM) are also provided.

Original languageEnglish (US)
Article number9034483
Pages (from-to)5626-5635
Number of pages10
JournalIEEE Transactions on Antennas and Propagation
Volume68
Issue number7
DOIs
StatePublished - Jul 2020

Keywords

  • Autoencoder (AE)
  • convolutional neural net
  • deep learning
  • microwave imaging
  • scattered fields

ASJC Scopus subject areas

  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Microwave Imaging by Deep Learning Network: Feasibility and Training Method'. Together they form a unique fingerprint.

Cite this