Compressive sensing-based correlation plenoptic imaging

Abstract

Correlation Plenoptic Imaging (CPI) is an innovative approach to plenoptic imaging that tackles the inherent trade-off between image resolution and depth of field. By exploiting the intensity correlations that characterize specific states of light, it extracts information of the captured light direction, enabling the reconstruction of images with increased depth of field while preserving resolution. We describe a novel reconstruction algorithm, relying on compressive sensing (CS) techniques based on the discrete cosine transform and on gradients, used in order to reconstruct CPI images with a reduced number of frames. We validate the algorithm using simulated data and demonstrate that CS-based reconstruction techniques can achieve high-quality images with smaller acquisition times, thus facilitating the practical application of CPI.