Phaseless Terahertz Coded-Aperture Imaging Based on Deep Generative Neural Network

Fengjiao Gan,Hongqiang Wang,Chenggao Luo,Ziyang Yuan

doi:10.3390/rs13040671

Fengjiao Gan, Hongqiang Wang + Show 2 more

Open Access

https://doi.org/10.3390/rs13040671

Copy DOI

Journal: Remote sensing	Publication Date: Feb 13, 2021
Citations: 1	License type: CC BY 4.0

Affiliation: National University of Defense Technology

Abstract

As a promising terahertz radar imaging technology, phaseless terahertz coded-aperture imaging (PL-TCAI) has many advantages such as simple system structure, forward-looking imaging and staring imaging and so forth. However, it is very difficult to recover a target only from its intensity measurements. Although some methods have been proposed to deal with this problem, they require a large number of intensity measurements for both sparse and extended target reconstruction. In this work, we propose a method for PL-TCAI by modeling target scattering coefficient as being in the range of a generative model. Theoretically, we analyze and model the system structure, derive the matrix imaging equation, and then study the deep phase retrieval algorithm. Numerical tests based on different generative models show that the targets with the different spareness can achieve high resolution reconstruction when the number of intensity measurements are smaller than the number of target grids. Also, we find that the proposed method has good anti-noise and stability.

Highlights

Terahertz Coded-Aperture ImagingOriginating from optical coded-aperture imaging [1,2] and radar coincidence imaging [3,4,5], terahertz coded-aperture imaging (TCAI) exploits aperture coded antenna [6,7]to modulate the THz wave randomly, and to achieve spatiotemporal independent wave distribution
Numerical tests based on fully-connected generator and deep convolutional generative adversarial network (DCGAN) [28] demonstrate the superiority of the presented method, which can achieve a high recovery rate for both sparse targets and extended targets when the intensity measurements are less than the number of target grids
The generative model follows the DCGAN framework that was trained on the TCAI-MNIST data set and trained on TCAI-FMNIST data set

Summary

Introduction

Terahertz Coded-Aperture ImagingOriginating from optical coded-aperture imaging [1,2] and radar coincidence imaging [3,4,5], terahertz coded-aperture imaging (TCAI) exploits aperture coded antenna [6,7]to modulate the THz wave randomly, and to achieve spatiotemporal independent wave distribution. Originating from optical coded-aperture imaging [1,2] and radar coincidence imaging [3,4,5], terahertz coded-aperture imaging (TCAI) exploits aperture coded antenna [6,7]. The target scattering coefficients can be solved through computational imaging method [8]. TCAI has many significant advantages such as high resolution, high frame-rate, all-time functionality. The forward-looking and staring imaging capability can be obtained without relying on any relative motion between radar and target [9], which is different from synthetic aperture radar or inverse synthetic aperture radar. The TCAI technique has potential applications in nondestructive detection [10], anti-terrorism checks[11,12], terminal guidance [9] etc

Results

Discussion

Conclusion