Large-aperture computational single-sensor microwave imager using 1-bit programmable coding metasurface at single frequency

📝 Abstract

The microwave imaging based on inverse scattering strategy holds important promising in the science, engineering, and military applications. Here we present a compressed-sensing (CS) inspired large- aperture computational single-sensor imager using 1-bit programmable coding metasurface for efficient microwave imaging, which is an instance of the coded aperture imaging system. However, unlike a conventional coded aperture imager where elements on random mask are manipulated in the pixel-wised manner, the controllable elements in the proposed scheme are encoded in a column-row-wised manner. As a consequence, this single-sensor imager has a reduced data-acquisition time with improved obtainable temporal and spatial resolutions. Besides, we demonstrate that the proposed computational single-shot imager has a theoretical guarantee on the successful recovery of a sparse or compressible object from its reduced measurements by solving a sparsity-regularized convex optimization problem, which is comparable to that by the conventional pixel-wise coded imaging system. The excellent performance of the proposed imager is validated by both numerical simulations and experiments for the high-resolution microwave imaging.

💡 Analysis

The microwave imaging based on inverse scattering strategy holds important promising in the science, engineering, and military applications. Here we present a compressed-sensing (CS) inspired large- aperture computational single-sensor imager using 1-bit programmable coding metasurface for efficient microwave imaging, which is an instance of the coded aperture imaging system. However, unlike a conventional coded aperture imager where elements on random mask are manipulated in the pixel-wised manner, the controllable elements in the proposed scheme are encoded in a column-row-wised manner. As a consequence, this single-sensor imager has a reduced data-acquisition time with improved obtainable temporal and spatial resolutions. Besides, we demonstrate that the proposed computational single-shot imager has a theoretical guarantee on the successful recovery of a sparse or compressible object from its reduced measurements by solving a sparsity-regularized convex optimization problem, which is comparable to that by the conventional pixel-wise coded imaging system. The excellent performance of the proposed imager is validated by both numerical simulations and experiments for the high-resolution microwave imaging.

📄 Content

REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <

1  Abstract—The microwave imaging based on inverse scattering strategy holds important promising in the science, engineering, and military applications. Here we present a compressed-sensing (CS) inspired large- aperture computational single-sensor imager using 1-bit programmable coding metasurface for efficient microwave imaging, which is an instance of the coded aperture imaging system. However, unlike a conventional coded aperture imager where elements on random mask are manipulated in the pixel-wised manner, the controllable elements in the proposed scheme are encoded in a column-row-wised manner. As a consequence, this single-sensor imager has a reduced data-acquisition time with improved obtainable temporal and spatial resolutions. Besides, we demonstrate that the proposed computational single-shot imager has a theoretical guarantee on the successful recovery of a sparse or compressible object from its reduced measurements by solving a sparsity-regularized convex optimization problem, which is comparable to that by the conventional pixel-wise coded imaging system. The excellent performance of the proposed imager is validated by both numerical simulations and experiments for the high-resolution microwave imaging.

I. INTRODUCTION

icrowave imaging is an important and powerful technique in science, engineering, and military.[1–3] In the context of high-frame-rate electromagnetic (EM) imaging, the obtainable temporal and spatial resolutions are mainly limited by the sustainable throughput of the imager’s memory, exposure time, and illumination conditions. Over the past decade, the coded aperture imaging system in combination with the sparsity-regularized reconstruction algorithm has gained intensive attentions.[4–8] A coded aperture imaging architecture relies on the use of a sequence of random masks, through which the modulated information of the probed object was fully captured by a single fixed sensor. Then, the information of the probed object can be faithfully retrieved with reduced number of measurements by solving a tractable optimization problem. When the probed object allows for a low-dimensional representation in certain transformed domain, either pre-specified or trained, such as DCT, wavelet, etc., it is known that such a single-sensor imager benefits from a fundamental fact that the number of measurements could be drastically reduced compared to that required by conventional imaging techniques. Interestingly, the required measurements

could be significantly less than the unknowns to be reconstructed, as claimed by the compressed sensing (CS) theory (e. g., Refs. [9–11]). In this area, a pioneering work is the well-known single-pixel camera invented in Rice University.[4] Basically, the working principle of the single-sensor imager for high-resolution imaging is described as follows. The wavefronts scattered from the probed object are firstly modulated by the random masks. Then, the modulated information is captured by a fixed single sensor. Finally, the information of the probed object is retrieved by a sparsity-regularized reconstruction algorithm. An essential issue of the single-sensor imager is the construction of multiple- mode modulators or masks, which is not well tackled and remains challenging in designing the controllable masks, especially in the microwave frequencies.
Metasurfaces have shown great promise in manipulating electromagnetic waves, including the microwave and millimeter waves in a flexible way, as evidenced by a number of interesting applications, such as ultrathin flat lens,[12–16] analogy signal processing,[17] high-resolution hologram,[18–20] and some other functional devices.[21–23] Therefore, metasurfaces have increasing abilities in designing the spatial modulator, and have become an important component of the cutting-edge imaging systems, especially for the single-sensor imager. More recently, the programmable coding metasurfaces [24,25] have been introduced to dynamically manipulate the EM waves in both microwave frequency and beyond.
Here, we propose a new method to realize the large-aperture single-sensor microwave imager by utilizing the 1-bit programmable coding metasurface composed of an array of voltage-controllable particles. Each metasurface particle illuminated by the incident wave could be in a state of two distinct responses: “1” for important radiation when the loaded voltage is at a high level, and “0” for almost negligible radiation. In this way, a sequence of different quasi-random radiation patterns are easily generated by managing the applied voltages of the coding metasurface, which could provide adequate modes in our imaging system. Compared with the systems of transforming frequency-dependent masks using dispersive and resonant metasurfaces,

This content is AI-processed based on ArXiv data.