Influence Of Noise On Reconstruction Of Sample Details With Coherent Diffraction Imaging

The achievable resolution of optical imaging is limited by the diffraction limit of the lens and aberrations.To overcome this limitation,lensless imaging techniques that get rid of optical focusing devices have been proposed.Coherent diffraction imaging is an ultra-high-resolution lensless imaging method that is able to restore complete optical wavefield information from the diffraction intensity information of an object.By making use of data redundancy,coherent diffraction imaging reconstructs the wavefield of the light wave and the complex amplitude of the sample by means of iterative phase recovery algorithms.With the development of ultra-short light sources and related algorithms,coherent diffraction imaging has developed rapidly and has a wide range of applications in nano-imaging materials,biomedicine,physics,and optoelectronic engineering.Traditional coherent diffraction imaging is limited by isolated samples.The generation of ptychography overcomes this defect.This method performs overlapping scanning of samples,providing high information redundancy and fast convergence,which has always been an important research technique for coherent diffraction imaging.Theoretically,the resolution of coherent diffraction imaging is only limited by the optical wavelength,but the existence of noise in experiments will affect the attainable resolution.Based on the ptychography technique,this thesis investigates the effect of noise on the reconstruction of sample details.The main research contents of this thesis are summarized as follows:1.The theoretical basis of coherent diffraction imaging is expounded.Firstly,we introduce the scalar diffraction theory of light waves,Fourier transform properties of lenses,and the reason why diffractive imaging systems can replace lens imaging.Finally,the related theory of phase recovery is introduced: oversampling rate,the flow of phase recovery algorithm and the mathematical basis of phase recovery algorithm-deconvolution filter and incremental Wiener filter in the principle of signal recovery.2.We introduce the principle of ptychographic coherent diffractive imaging,and performed simulations by programing the related algorithm code.The sample image was reconstructed with the programmed algorithm,and several simulation experiments were carried out to explore the influence of the overlapping rate of scanning on the reconstruction effect by changing the step size of the scanning.It was found that the reconstruction effect was not always proportional to the overlapping rate.The probe overlapping rate of about 87.5% corresponds to the best reconstruction effect with the same iteration times.3.Based on the algorithm of ptychographic coherent diffractive imaging,we added Poisson noise and Gaussian noise to the diffractogram of the sample respectively,and investigated their influence on the reconstruction of sample transmittance details.The results show that the reconstruction of sample details with low transmittance were mostly affected by the noise.When the signal-to-noise ratio is lower than 6,the sample details with different transmittances cannot be reconstructed.When the signal-to-noise ratio is greater than 22,noise has no influence on the reconstruction of sample details with different transmittances.