Research On Noise Reduction Technology Of X-ray Grating Phase-contrast Imaging

X-ray imaging technology used in the fields of clinical medicine and nondestructive testing,in order to realize the detection of light elements,X-ray phase contrast imaging is proposed.The research in the past two decades has rapidly developed,and the X-ray phase-contrast imaging based on the Talbot-Lau effect has made it free from the limitations of synchrotron radiation sources.As the most clinical application prospect of phase contrast imaging method,the important task of grating phase contrast imaging technology is to provide high-quality images for clinical diagnosis.Therefore,it is necessary to study image processing methods suitable for raster imaging technology.In the actual image acquisition process,due to equipment defects and external environmental disturbances,the finally acquired image will be subject to noise pollution,streaks,etc.,which will seriously degrade image quality and affect the resolution of image details.In order to solve the above problems,we designed and built an X-ray grating imaging system.The system parameters were adjusted to minimize the noise of the system,and the final image was denoised.Based on the X-ray grating phase-contrast imaging system,this dissertation mainly focuses on the research of detector calibration,experimental platform construction and image noise removal methods in the laboratory grating imaging system.Firstly,based on the phase-contrast imaging principle of X-ray gratings,the noise components in the system are analyzed,the optimal sensitivity formula of the system is deduced,and the system length and correlation that can obtain the best sensitivity under the conditions of the experimental platform are determined.parameter.Secondly,the detector used in this experimental system is a CMOS flat panel detector.The noise caused by the CMOS sensor itself is analyzed.The noise is mainly divided into random noise and three kinds of inherent noises.Correspondingly,the causes and characteristics of each noise are formulated accordingly.The algorithm performs noise reduction and the stripe noise in the detector is removed by Fourier transform andself-correlation interpolation respectively.Thirdly,from the perspective of the X-ray source,in order to control the amount of radiation,three factors related to the amount of radiation are analyzed: the effects of tube voltage,tube current and exposure time on image noise,and these three indicators are determined.The optimal parameter values in this experimental system;the image was studied when the random noise in the experimental system was eliminated using the frame integral averaging method the problem of selecting the best number of superimposed frames;in the process of comparing image quality,since the average values of the images are different,the coefficient of variation is introduced to compare how much noise is in the two images.Fourthly,due to the defects of the grating fabrication,a new method is proposed to eliminate the noise caused by the grating defects in the imaging process of the polyethylene cylinder,thereby assisting the experimental researchers in analyzing the advantages and disadvantages of the grating imaging system.