Font Size: a A A

Research On Image Reconstruction Algorithms And Design Of Software For Multiphase Flow CT System

Posted on:2011-01-14Degree:MasterType:Thesis
Country:ChinaCandidate:Q XueFull Text:PDF
GTID:2250330392969814Subject:Detection Technology and Automation
Abstract/Summary:PDF Full Text Request
Multiphase flow is prevalent in petroleum refining, chemical industry,biotechnology and pollution control. High-precision, real-time multiphase flowmeasurement is of great significance to the control and management of industrialproduction. Owning to the advantages of non-invasiveness, visualization andhigh-precision, computed tomography (CT) technique is widely applied to clinicalmonitoring and multiphase flow measurement.In recent years, positive evolutions have been made in industrial CT technique.While compared with medical CT, the CT applied in process engineering is of thelimitation in the number of projections measured, and consequently in the imagingquality as the inverse problem is seriously ill-posed. Moreover, imaging software haslagged behind hardware design and reconstruction algorithm as researchers paid moreattention to the latter two, both the real-time performance and display effect areunsatisfactory. In response to the above problems, thorough researches have beenconducted mainly on image reconstruction algorithms and software design for CTsystem in this thesis. The main work and results are as follows.1. Research on Image Reconstruction Algorithms for CTCommon algorithms for image reconstruction from projections have beensummerized and applied to multiphase flow γ-CT. Aiming to improve space resolutionwith limited projection data, two approaches of expanding projection matrix weredeveloped, thereafter implemented image reconstruction based on variousreconstruction algorithms. To reconstruct better edge and realize multipledimensioned dissection, we triangulated the imaging domain based on Delaunaytriangulation technique, then calculated projection coefficient of the triangular cellwith the normalized distance between the image of the triangle‘s gravity center and itsnearest detector. By improving the performance of reconstruction algorithm with levelset method and image processing techniques, the number of iterations needed forobtaining satisfactory images is largely reduced, while the image precision isobviously enhanced. Both simulation and experimental results show that our CTsystem can reconstruct satisfactory cross-sectional images from limited projections. The reconstructed results will contribute to flow regime identification or characteristicparameter extraction in industrial multiphase flow measurement.2. Design of Software for CT SystemA CT software system named TJU_CT_III has been designed based on32-bitsWindows platform and Visual C++programming environment. It is composed of thefollowing function modules: USB communication module; algorithms applicationmodule; display module; data saving and replaying module. It realizes fast dataacquisition based on USB communication. Various algorithms are integrated in thealgorithms application module. The display effect is largely improved utilizingOpenGL. Moreover, this program can save and replay experimental data.
Keywords/Search Tags:Computed Tomography, Image Reconstruction Algorithm, Projection Matrix, USB Communication, OpenGL
PDF Full Text Request
Related items