Application Of Kalman Filter In Image Reconstruction Algorithm Of Electromagnetic Tomography

Electromagnetic tomography (EMT) is one new kind of the electricaltomography technologies. It calculates spatial distribution of the electricallyconductive and magnetically permeable materials through electromagnetic sensorarray around detection pipeline. Electromagnetic tomography technique isnon-invasive, non-contact, detecting magnetic materials, etc, and so it has a broadapplication prospect in the field of industrial process control, biomedical research,foreign body monitoring. EMT is still in the stage of laboratory research, how toapply the EMT to actual production and living as soon as possible is the researchobjectives of the researchers. In this paper, starting from the basic principles ofelectromagnetic tomography, using Comsol software for3-D simulation of the EMTsystem, and proposing the EMT image reconstruction algorithm based on Kalmanfilter.In this thesis, completed the following work: First of all, through access to alarge number of domestic and foreign literature, summarized Electromagneticdetection technology development process, application status, and imagereconstruction algorithm. Secondly, elaborated structure and working principle of theEMT system basis, started from the Maxwell equations to derive mathematical modelof EMT; and with finite element software Comsol established a the EMT system of3-D simulation model, studyed simulation of EMT positive and the sensitive fieldcharacteristics of the empty field and flow pattern affect to sensitive field. Usedparameter extraction method to calculate the conductivity and magnetic permeabilityof the sensitivity of the matrix, and provided a priori information for the imagereconstruction of the EMT. Finally, studied the inverse problem of electromagnetictomography, that was, the image reconstruction algorithm. For image reconstructionalgorithm for imaging accuracy was not high. This paper presented an EMT imagereconstruction algorithm based on Kalman filter. The basic principle of the Kalmanfilter introduction derived the mathematical model of the Kalman filter image reconstruction algorithm. In this method, linear back projection (LBP) imaging resultswas the initial value, and then used the Kalman filter method for image reconstruction,finally, and did a simulation experiment, and compared with Landweber iterativemethod of image reconstruction results. Simulation results showed that the proposedalgorithm improved image reconstruction error and correlation coefficient and otherindicators than Landweber iterative method. This also verified that the proposedalgorithm was effective.