Imaging and positioning through scattering media noninvasively by bi-directional exposure

Abstract

Speckle autocorrelation is a popular method for imaging through scattering media. Currently, most proposed methods in literature can only recognize the pattern and the relative movement ratio of a hidden object. Position and size quantification, the other content of imaging, is just as important as object recognition. Here, we propose a non-invasive method to estimate the axial position and size of a hidden object based on speckle autocorrelation imaging. A unique geometry setup for bi-directional exposure is applied to extract the position and size information. The image of the object can be reconstructed from the recorded speckle pattern via phase retrieval techniques. According to the geometry relationship, we calculate the size and axial position of the object, and further, obtain the lateral or axial displacements when the object is shifted in both simulation and experiment. This method opens a door for noninvasive imaging and positioning through scattering media, will impact the fields of biomedical imaging, object tracking, etc.