Deep Learning for Sensing Matrix Prediction in Computational Microwave Imaging With Coded-Apertures
Jiaming Zhang, Rahul Sharma, María García-Fernández, Guillermo Álvarez-Narciandi, Muhammad Ali Babar Abbasi, Okan Yurduseven
Abstract
This work aims to simplify the characterization process of coded-apertures for computational imaging (CI) at microwave frequencies. A major benefit of the presented technique is the minimization of the processing time needed to calculate the system sensing matrix for microwave CI-based compressive sensing applications. To achieve this, a deep learning-based approach is proposed which is capable of generating the sensing matrix using features learned directly from the coded-aperture distribution. To avoid the vanishing gradient problem, the proposed deep learning network contains skip connections. On 1,000 testing samples, the average normalized mean-squared-error (NMSE) calculated between the sensing matrix generated by the conventional method and that predicted by the proposed network is 0.0036. Moreover, the average mean-squared- error (MSE) calculated between the images reconstructed using the conventional and predicted sensing matrix is 0.00297. In addition to providing high-fidelity estimations with minimized error, we demonstrate that using the trained network, the prediction of the sensing matrix can be achieved in 0.212 s, corresponding to a reduction of 65% in the computation time needed to calculate the sensing matrix. This has significant outcomes in achieving real-time operation of CI-based microwave imaging systems.