Experimental Research On Compressive Sensing Ghost Imaging

Compressive sensing is used in ghost imaging technology in this dissertation to process data sampling and image reconstruction in order to reduce the measurement times and improve the quality of reconstructed images.First, the development history, theory model and experimental scheme of the ghost imaging are introduced in this dissertation in detail. The experimental principle and characteristics of optical source are discussed especially of entangled light ghost imaging system, thermal light ghost imaging system, pseudo-thermal light ghost imaging system.Secondly, compressive sensing is introduced, including the theory model and the sparsity or compressibility, linear observation and reconstruction algorithm. The main operation process of the existing compressive sensing algorithm is described. The imaging results of 5 classical reconstruction algorithms are researched by simulation.Thirdly, the compressive sensing ghost image system is researched by simulation and experiment with the discrete cosine transform matrix as the sparse matrix. The experimental results of the Orthogonal Matching Pursuit algorithm, Subspace Matching Pursuit algorithm and Iterative Reweighted Least Square algorithm show the decrease of measurement times and the increase of signal-to-noise rate of the reconstructed images.Finally, based on the two order compressive sensing ghost imaging, two kinds of high order compressive sensing ghost imaging schemes are proposed. The simulation and experimental results show that the signal-to-noise rates of the two kinds of high order compressive sensing ghost imaging systems are better than two order of ghost imaging.