Image Imputation: A Mmwave 3D SAR Imaging Application

Image imputation plays a critical role in ensuring high-quality image reconstruction in circumstances including the occurrence of missing data points, such as during mmWave 3D Synthetic Aperture Radar (SAR) imaging. In this study, we tackle the issue of missing data points in mmWave SAR imaging by employing various imputation techniques to restore the missing image data matrix. We employed several image quality metrics to quantify and evaluate the image quality of the restored images and assess and compare the performance of different imputation methods. We applied the image imputation techniques and evaluated the image quality metric for SAR data collected in two different settings: a controlled environment, which is an anechoic chamber, and a noisy environment that is typical of a laboratory setting. Our investigation demonstrated that the efficacy of various image imputation techniques, with regard to image quality metrics, varied across different scenarios and was contingent upon the quantity of missing data points. This illustrates the necessity of choosing suitable imputation algorithms tailored to the unique characteristics of the image data and the surrounding environment.