Depth-resolved and auto-focus imaging through scattering layer with wavelength compensation.

Abstract

Imaging techniques through turbid materials have been extensively studied in recent years. The challenge now is to recover objects in a large field of view with depth-resolving ability. We present a method to image through a thin scattering layer automatically with the depth of the object detectable. By revealing the wavelength-depth-matching relation based on the axial memory effect, this method can automatically search the optimal wavelength of the reference light and compute the depth of the object. The no-reference image quality assessment function and rule-based searching algorithm are used in the searching process. The proposed method is promising for dynamic object tracking.