Research On Denoising Methods Of Ghost Imaging

Ghost imaging,also known as correlation imaging,is a new method to obtain imagings through quantum entanglement or light field fluctuation.Different from the traditional imaging method,ghost imaging can reconstruct the imaging of the unkown object.In recent years,ghost imaging has become one of the hotspots in the field of quantum optics.In this thesis,we study the de-noising methods in ghost imaging.(1)For the influence of random noise on ghost imaging system,we propose a denoising method based on pre-emphasis and four-step phase shift,which is called pre-emphasis four step phase shift ghost imaging denoising method.In the method,the intensity of the speckle pattern in ghost imaging system is firstly enlarged with pre-emphasis technique,so as to be far greater than that caused by random noise.Before imaging reconstruction,the pre-emphasis is removed from the bucket detector value.And a clearer image is finally recovered with the four-step phase shift reconstruction.The simulation results verify the feasibility of the proposed method.By comparing the peak signal-to-noise ratio(PSNR)with and without pre-emphasis technique in different random environment noise,it can be found that the proposed method can effectively reduce the influence of system's random noise on ghost imaging even if the random noise is a big one.The simulation results also show that the proposed method can reduce the influence of noise on the image quality when the image is compressed.(2)Considering that mean filtering is one of the common image denoising methods,we proposes a ghost imaging denoising method based on mean filtering.In this method,the shifted speckle patterns with different directions are first designed according to the mean filtering template,then the shifted speckle patterns are illuminated on the object individually.The detection values with different direction shifted speckle patterns are obtained from ghost imaging system.By the superposition of the bucket detection value in each direction,the relationship between the bucket detection values and the shifted speckle patterns is achieved.With the second-order correlation,the imaging of the unknown object is reconstructed.The simulation results show that the influence of noise on the imaging quality is effectively reduced by the proposed method,especially for the Gaussian white noise and the salt and pepper noise.In addition,the proposed method is suitable for both intensity and phase objects.