Lensless multi-core-fiber-based imaging from speckle autocorrelation under highly coherent illumination

Flexible fiber-optic lensless endoscopic imaging plays a significant role in medical and biological applications. Lensless imaging based on far-field speckle of multi-core fibers under spatially incoherent light source illumination relied on the angular optical memory effect, resulting in limited field of view and susceptibility to interference effects. Image reconstruction from fiber speckle intensity under the highly coherent light illumination has been investigated using classical coherent diffraction theory, but it imposes limitations on imaging distance and detection positions. However, existing alternative solutions involve reducing coherence level to satisfy the angular optical memory effect, which increases system complexity and sacrifices the advantages of single-frame imaging. The reconstruction of intensity-based images remains a significant challenge under highly coherent illumination, primarily due to the interference effect. In this study, an imaging algorithm that reconstructs the original image solely from the intensity of far-field fiber speckle under coherent illumination conditions is proposed. The proposed method is based on statistical optics theory and aims to extract the relevant information within coherent speckle patterns to reconstruct the image. Our approach expands the imaging field of view, simplifies the imaging setup, and broadens the range of viable light sources for speckle correlation imaging. The effectiveness of the algorithm is verified by simulated experiments and a real imaging system.