Expanding the memory effect in scattering imaging by manipulating photon distributions

Speckle correlation techniques based on the memory effect can recover objects from multiple scattered light without the information of the scattering media. The existing methods, however, are limited by the inherent angular width of the memory effect, and the retrieval algorithms frequently converge to erroneous solutions. In this paper, we propose an autoshrinking algorithm to expand the field of view in scattering imaging. We also develop a pretreatment iteration scheme to achieve high accuracy reconstruction. The autoshrinking method improves the quality of the energy spectrum by adaptively shrinking the speckle autocorrelation to its expanded central peak region and enhancing the ratio of the ballistic photons to the scattered photons. In the pretreatment iteration, the Brenner gradient function of a short testing iteration is employed to select the initial random Fourier phase to ensure the iteration can quickly and accurately converge. Experiments and simulations successfully confirm the effectiveness of our method.