Phase Imaging through Scattering Media Using Incoherent Light Source

Phase imaging normally employs coherent a light source while an incoherent is not preferred due to its random wavefront. Another challenge for practical phase through scattering media, which scatter the photons in manner and lead seriously distorted images of speckles. Based on convolutional neural network (CNN), this paper presents approach media using source. A CNN was trained utilized reconstruct target from captured Similarities over 90% between reconstructed their have been achieved. It concluded that can be used as illumination with assistance deep learning technology. This record refractive indices transparent samples, might application biomedical imaging.