Research On Image Quality Enhancement Method In Micro-computed Tomography Reconstruction

Computed Tomography(CT)has the advantage of being able to get the internal structure information of the object without damaging the object.It plays an important role in clinical examination,industrial device damage detection,material composition analysis,and scientific research.The study of CT can be divided into improving CT reconstructed image quality and other improvement under the premise of CT reconstructed image quality.This thesis is devoted to improving the quality of reconstructed images from software algorithms based on the micro-CT imaging system in our laboratory.The main work can be divided into the following points:(1)In order to improve the resolution of reconstructed images,a framework of improving the micro-CT reconstructed images resolution based on single image super-resolution is proposed.This framework applies the existing single image super-resolution method based on the double-dictionary sparse reconstruction to the sinogram domain and the projection domain.And then obtain the resolution improved reconstructed image by filtering back projection algorithm.Finally,the simulation data and the actual measured data experiment are used to verify the resolution enhancement effect.(2)In order to improve the contrast of reconstructed images,a micro-CT ring artifact correction method based on polar coordinate transformation is proposed.Firstly,the reconstructed image in the Cartesian coordinate system is transformed into the polar coordinate system.And then convert the difference map between the low filtering of the original image and the original image to the polar coordinate system for the determination of the artifacts position with corresponding threshold.Next,the nearby normal pixels are selected for artifacts correction.Finally,the image is converted to the Cartesian coordinate system.Also the simulation data experiment and the real measured data experiment are carried out to verify the effect of ring artifact correction.(3)In order to improve the contrast of the reconstructed images and distinguish the soft tissue effectively.A dual energy CT reconstruction method based on NLM-MMD is formed by combining non-local mean(NLM)filtering algorithm and dual energy CT multi-material decomposition(MMD)algorithm.In this method,the reconstructed images under the two energy spectra are filtered by non-local mean algorithm first.And then the two energy spectra reconstructed images are decomposed to three best combinations of base materials base on the measured absorption coefficient of the base materials.Finally,the double energy CT decomposition effect is verified by the measured data experiment.(4)In order to comprehensively improve the quality of reconstructed images,the above three methods are combined and form a complex micro-CT reconstruction image quality improvement method.In this complex method,the resolution of the reconstructed image is improved first.And then the ring artifacts in the reconstructed image are corrected.Finally,the double-energy CT base material decomposition is carried out by NLM-MMD method.Also the effect of micro-CT reconstruction image quality enhancement in two-dimensional case is verified.