Font Size: a A A

Research On Reconstruction And Material Decomposition Techniques For Dual Energy X-ray Container CT

Posted on:2019-07-07Degree:DoctorType:Dissertation
Country:ChinaCandidate:T ZhaoFull Text:PDF
GTID:1362330590951491Subject:Nuclear Science and Technology
Abstract/Summary:PDF Full Text Request
Container freight is the main way of goods transportation in the current globalization economy.At present,the container radiography inspection system,which is the main detection measure for the customs of various countries,has some limitations.The image information has overlapping in the direction of ray,which leads to the occurrence of missed detection.Dual energy CT imaging technology can provide three-dimensional structure and material information of the detected objects,and greatly improve the accuracy of detection.However,compared to the low energy dual energy CT which has a lot of applications,the container high energy(MeV)dual energy CT has a great technical difficulty.At present,there are no such products,and the public research data are very few.This paper aims to study the container high energy dual-energy CT imaging technology,to solve the core technical difficulty of image reconstruction and material decomposition,and to lay the theoretical foundation for the development of the container high energy dual-energy CT product.In this paper,based on the interaction of high-energy x-ray photons and materials,an innovative decomposition model of double basis material without photoelectric effect is proposed for the container high energy dual-energy CT.Compared with the direct use of the double basis material decomposition model,the high energy x-ray attenuation coefficient function space can be more accurately expressed by the proposed model.On the basis of this model,a projection-domain decomposition method with photoelectric effect removal is proposed.Simulation experiments show that when the noise level of the projection is not too high,for the high atomic number material,the method can effectively improve the accuracy of dual energy CT material decomposition.In view of the variety of materials and the wide range of atomic numbers in the customs container inspection,a dynamic double basis material decomposition model is further proposed.The precision of the description of the high energy x-ray attenuation coefficient function space is further improved by the proposed model.Then a dynamic double basis material projection-domain decomposition method is proposed.Simulation experiments show that when the projection noise level is not too high,for the high atomic number material,the proposed method can further improve the accuracy of material decomposition ?When the density of container cargo is large,the noise of projection data will increase significantly.At this time,the decomposition result of the above projection-domain method will be greatly affected.In this paper,the dual-energy reconstruction image data are clustered according to the data density center.Based on this image-domain material clustering and the dynamic double basis material decomposition model without photoelectric effect,an innovative high energy dual-energy CT iterative decomposition method is proposed.The simulation experiments show that the material clustering based on the density center of the dual-energy reconstruction image data is more consistent with the actual material distribution,and the results of the iterative decomposition method not only keep the accuracy of the material information,but also have far higher image quality than the projection-domain decomposition results.In this paper,a container high energy dual-energy CT experimental system is built by using 9/6MeV accelerator and linear array detector.Experimental results show that the performance of methods in this thesis has reached the design requirements of dual energy CT and meets the application requirements.
Keywords/Search Tags:Container inspection, High Energy Dual-energy CT, Projection-domain decomposition, Image-domain decomposition, Iterative decomposition
PDF Full Text Request
Related items