• Chinese Optics Letters
  • Vol. 20, Issue 4, 041101 (2022)
Weihao Wang1, Xing Zhao1、2、*, Zhixiang Jiang1, and Ya Wen1
Author Affiliations
  • 1Institute of Modern Optics, Nankai University, Tianjin 300350, China
  • 2Tianjin Key Laboratory of Optoelectronic Sensor and Sensing Network Technology, Tianjin 300350, China
  • show less
    DOI: 10.3788/COL202220.041101 Cite this Article
    Weihao Wang, Xing Zhao, Zhixiang Jiang, Ya Wen. Deep learning-based scattering removal of light field imaging[J]. Chinese Optics Letters, 2022, 20(4): 041101 Copy Citation Text show less

    Abstract

    Light field imaging has shown significance in research fields for its high-temporal-resolution 3D imaging ability. However, in scenes of light field imaging through scattering, such as biological imaging in vivo and imaging in fog, the quality of 3D reconstruction will be severely reduced due to the scattering of the light field information. In this paper, we propose a deep learning-based method of scattering removal of light field imaging. In this method, a neural network, trained by simulation samples that are generated by light field imaging forward models with and without scattering, is utilized to remove the effect of scattering on light fields captured experimentally. With the deblurred light field and the scattering-free forward model, 3D reconstruction with high resolution and high contrast can be realized. We demonstrate the proposed method by using it to realize high-quality 3D reconstruction through a single scattering layer experimentally.

    1. Introduction

    Light field (LF) imaging technology has shown great significance in recent years for its high-temporal-resolution 3D imaging feature through simultaneously capturing the 2D spatial and 2D angular information of light [four-dimensional (4D) LF information][15]. Especially, the LF imaging method based on the wave-optics model and LF point spread function (LFPSF) allows 3D deconvolution for the high-quality single-shot volumetric reconstruction[6,7]. However, in some applications of LF imaging, a scattering medium is present in the scenes, such as biological tissue, fog, and turbid water[816]. In these imaging scenes, signal light could still be captured, and 3D reconstruction could be conducted, but scattered light produced by the scattering medium introduces blur and scattering background artifacts to the 3D reconstruction image, which lead to low resolution and low contrast. For the LF imaging method based on 3D deconvolution, sequence recorded frames are utilized to extract ballistic light and undo the effect of scattering on the LF in real-time localization of neuronal activity[10,11], but the demand of multiple frames rather than single-shot reduced the flexibility of the method. Signal light and scattered light are separated and reconstructed separately[14], but this will increase the amount of calculation and the solving difficulty. In a general method, scattering is incorporated into the LF imaging forward model to deal with the case of mismatching between the scattering-free model and the LF with scattering, and then the scattering background artifacts of 3D reconstruction can be removed[8,9]. However, scattering introduces degradation of the LF image and produces blur patterns of LFPSF. These make the inverse problem of the 3D deconvolution more ill-conditioned and lead to a noise-sensitive result[17].

    Copy Citation Text
    Weihao Wang, Xing Zhao, Zhixiang Jiang, Ya Wen. Deep learning-based scattering removal of light field imaging[J]. Chinese Optics Letters, 2022, 20(4): 041101
    Download Citation