Research On Fast Ghost Imaging Based On Binary Speckle

Since the ghost imaging technology was proposed,it has attracted widespread attention due to its strong anti-noise capability and high resolution.However,the application of traditional ghost imaging has been limited due to its long imaging time.Although the introduction of schemes such as binary speckle ghost imaging has greatly shortened the imaging time and imaging quality,people are still looking for the scheme that takes fewer measurements to restore the image.We focuses on the realization of fast ghost imaging in the dissertation.The study addresses the problem of long imaging time of current ghost imaging systems,and proposes a scheme that invokes a compressed perceptual reconstruction algorithm based on the use of binary scattered illumination patterns.Among them,the selected compression-aware reconstruction algorithms are the subspace pursuit algorithm,the orthogonal matching pursuit algorithm,and the total variation minimization by augmented lagrangian and alternating direction algorithms.The scheme better meets the demand for high quality and speed of fast ghost imaging.The main research of this paper are as follows:1.The ghost imaging comparison experiments under different illumination patterns are carried out.The results show that compared with the ghost imaging under random illumination patterns,using binary speckle as the illumination pattern for ghost imaging can effectively reduce the number of measurements and shorten the imaging time.2.The numerical simulation and experimental verification of incomplete sampling Hadamard ghost imaging based on different compressed sensing reconstruction algorithms are carried out,and the comparison results under the quantitative evaluation criteria are obtained by measuring the imaging time,calculating the image evaluation index and other physical quantities.The final experimental results show that: When reconstructing binary images,the subspace tracking algorithm combined with incomplete sampling Hadamard ghost imaging can achieve better image reconstruction results at low sampling rates.When reconstructing grayscale images,the combination of orthogonal matching pursuit algorithm and incomplete sampling Hadamard ghost imaging can achieve a clearer reconstructed image on the basis of fast imaging.3.The numerical simulation and experimental verification of compressed sensing Hadamard ghost imaging based on different resolutions are designed.The research results show that the combination of different reconstruction algorithms and sub-sampling Hadamard ghost imaging for different image reconstruction demands can reconstruct high-quality images in a short time.4.The sampling problem of reconstructing the grayscale test object under high resolution is experimentally explored,and the results show that a clear original image of the test object can be reconstructed by selecting the sampling rate within an appropriate range.The research results of this paper not only provide a new idea for fast ghost imaging,but also lay a foundation for the application of ghost imaging technology in detecting target object information,and further promote the practical process of ghost imaging technology.