Resolution Enhancement Based on Antenna Aperture Compensation in Synthetic Aperture Radar Imaging With Hyper-Laplacian Prior Deconvolution

This article presents the results of an algorithm for resolution enhancement, which aims at compensating the influence of finite aperture. As a significant technical index, resolution determines the imaging ability and recognition rate of relevant applications. However, the large antenna aperture influences the imaging quality. To compensate the influence of antenna aperture, a deconvolution process of the antenna aperture is required. Moreover, there are two important elements: deconvolution technique and deconvolution kernel. Deconvolution kernel is related to the aperture field of the applied horn antenna. Compared with analytical solution, full-wave calculations of the aperture field estimation are proved as a superior method. Whereas hyper-Laplacian prior deconvolution is addressed, because the prior consideration of image distribution recovers better details than Wiener filtering. Through experimental validation, the imaging resolution has been improved up to 1/6 of the aperture size, which is three times better than the classical limit without loss of signal-to-noise ratio (SNR). Essentially, this approach is not limited by frequencies. As an advanced imaging technique, our proposed method can be applied in both gigahertz (GHz) and terahertz (THz) imaging with an outstanding resolution. Tremendous applications, such as security inspection, remote sensing, nondestructive testing (NDT), and medical imaging, can benefit from it. Our proposed resolution enhancement algorithm also benefits target recognition rate and measuring accuracy in imaging applications.