Prior-image-filtered ghost imaging

In this study, a novel precise reconstruction method was proposed for ghost imaging. In traditional ghost imaging (TGI), image quality deteriorates in proportion to the ℓ² norm of the observed object. However, the proposed method reduces the effective ℓ² norm by filtering an unknown direct current component and an arbitrary alternating current component derived from a pre-measured rough image. Mathematical analysis demonstrated that the proposed method outperformed TGI in terms of precision. Numerical simulations and experiments validated that the mean squared error in the proposed method was considerably lower than that in existing online algorithms, demonstrating its effectiveness for high-resolution single-pixel imaging.