We present a new color computational ghost imaging strategy using a sole single-pixel detector and training by simulated dataset, which can eliminate the actual workload of acquiring experimental training datasets and reduce the sampling times for imaging experiments. First, the relative responsibility of the color computational ghost imaging device to different color channels is experimentally detected, and then enough data sets are simulated for training the neural network based on the response value. Because the simulation process is much simpler than the actual experiment, and the training set can be almost unlimited, the trained network model has good generalization. In the experiment with a sampling rate of only 4.1%, the trained neural network model can still recover the image information from the blurry ghost image, correct the color distortion of the image, and get a better reconstruction result. In addition, with the increase in the sampling rate, the details and color characteristics of the reconstruction result become better and better. Feasibility and stability of the proposed method have been verified by the reconstruction results of the trained network model on the color objects of different complexities.
An encryption method based on computational ghost imaging (CGI) with chaotic mapping and DNA encoding is proposed. To reduce the amount of keys in the CGI-based encryption system, the chaotic mapping algorithm is used to generate the random sequence as the speckle measurement matrix of CGI system. The measurement data of the bucket detector is subjected to block and DNA operations, which introduce the nonlinear characteristics in the encryption process. The problem of linear vulnerability of the encryption system has been greatly improved. Numerical simulation results show that, compared with the traditional CGI-based encryption method, the proposed method greatly reduces the amount of keys, increases the key space and enhances the security of the system.
Fourier-domain full-field optical coherence tomography (FD-FF-OCT) has the advantages of high resolution and parallel detection. However, using parallel detection can result in optical crosstalk. Toward minimizing crosstalk, we implemented a very fast deformable membrane (DM) that introduces random phase illumination, which can effectively reduce the crosstalk by washing out fringes originating from multiply scattered light. However, for one thing, although the application of DM has reduced the crosstalk problem in parallel detection to a certain extent, there will still be a lot of background noises, which may come from the circadian rhythm of the sample and multiple scattered photons. The problem could be solved by employing the adaptive singular value decomposition (SVD) filtering. We also combined SVD with the cumulative sum method, which can improve image resolution well. For the other thing, the random phase introduced by DM in the spectral domain will cause axial crosstalk after inverse Fourier transform. As far as we know, we are the first team to notice axial crosstalk and proposes that this problem can be solved by controlling the deformation range of DM. We have carried out a theoretical analysis of the above methods and verified its feasibility by simulation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.