Memory Effect Based Filter to Improve Imaging Quality Through Scattering Layers

Abstract

Recent extensive studies have promoted the imaging techniques through scattering layers to a reality but the imaging quality still needs to be improved. We demonstrate a high-speed method to image objects hidden behind a thin scattering medium. By analyzing the imaging degradation model of the speckle correlation methods, a memory effect (ME) based filter is proposed to improve the image quality of reconstructed images with deconvolution operation. Comparing to Wiener filter and other related methods, the ME-based filter can optimize the recovered images with the smaller mean square error, suppression of noise background, and central maximum. The proposed ME filter is also applied to improve the imaging method proposed for dynamic imaging through scattering layer with a reference object. The artifact generated by the deviation of retrieved point spread function can be effectively eliminated with a better imaging quality. This ME-based filter is promising to improve the imaging quality for other deconvolution and speckle correlation methods.