The CT Image Reconstruction Method Study Based On Adaptive Wavelet Denoising

Computed tomography(CT)is a technique of using ray source having a certain energy to tomography and getting the two-dimensional tomographic images with the projection data obtained in the case without damaging the physical structure of the object.It has been widely used in medicine,machinery,aerospace,nuclear industry and other fields.CT is becoming one of the main diagnostic means.At the same time,people are concerned about the potential hazards of the human body caused by CT radiation dose in a diagnosis.With reducing of the projection angle and the number of detector can reduce the radiation dose,the reconstructed image is mixed with significant noise and led to serious degradation of the quality In case of incomplete data.Therefore,how to reduce the radiation dose while getting highly accurate,highly detailed image has become a hotspot in the research of CT.Two following methods are mainly studied in this thesis:Firstly,a local CT reconstruction wavelet-based algorithm transform is proposed to reduce the radiation dose to the target.It choose high-order vanishing moments wavelet transform to process projection data,use its local characteristics of compactly supported partial reconstruction.Setting suitable wavelet base function gets the local wavelet reconstruction algorithm based on the filtered back projection(FBP)algorithm.It can overcome the non-locality of FBP algorithm.Simulation and comparison show that this algorithm can get better results by reducing the projection data.Secondly,In this thesis,an adaptive combination wavelet-based denoising algorithm is proposed to the noise suppression.The noise will cause to serious degradation of image.Therefore the noise in the projection data should be effectively filtered.It's based on wavelet threshold denoising algorithm,use Bayesian estimation denoising algorithm to filter noise,by analyzing the noise statistical characteristics of in wavelet domain data.Then use them filtering with the Wiener filter designed to achieve the effect of combination denoising.Simulation and comparison show that noise suppression ability of the this algorithm in the projection data consistent conditions has been significantly improved.Finally,an adaptive local wavelet-based denoising algorithm is proposed.For the weakness of higher demand of data completeness to the combination denoising algorithm,a local reconstruction nature is introduced to get a clearer reconstructed image,by using aprojection data obtained by the combination denoising to the local reconstruction algorithm for completing reconstruction.Combined with the actual projection angle and properly adjusting the number of detectors in reducing radiation dose,while achieving better reconstruction results.