A method of 3D light field imaging through single layer of weak scattering media based on deep learning

摘要

The study of imaging through scattering media especially 3D imaging is of great significance in many fields such asbiomedical imaging. Recently, deep learning has been widely used in the field of information processing with itsremarkable performance. In this paper, we proposed a method of three - dimensional imaging through scattering mediabased on deep learning. This method uses the deep neural network to process the information captured by the light fieldimaging system based on the microlens array, recovering the no-scattering 4D light field information, and then realizethree-dimensional reconstruction by using the processed light field information. Deep learning method requires a largenumber of samples. But in many environments, it is difficult to obtain a large number of three-dimensional samples throughexperiment. To solve this crucial problem, we use incoherent light propagation model to simulate the light field propagationand generate samples which contains three-dimensional information through simulation. In this paper, we simulated thepropagation of radiation emitted from objects behind a single layer of weak scattering media, generated a large number ofsamples of 4D light field information by simulation, trained the neural network and processed the test data set generatedby simulation, and we realized the deblurring of the light field information which contains information of multiple layersof flat semitransparent objects, which could be used to realize the 3D reconstruction.