Study On Inverse Problem Solving And Image Fusion Method For Electrical Capacitance Tomography

Electrical capacitance tomography (ECT) as an important field of theprocess tomography grows rapidly in recent years. Dependent on the workingprinciple of capacitance measurement, electrical capacitance tomography hasmany advantages such as safe, non-invasive detection, wide application range andso on, based on this, electrical capacitance tomography is used in many fields.Because of the limitations of ECT systerm, the precision of image reconstructionin electrical capacitance tomography system has been greatly restricted. Based onthe electrical capacitance tomography system with12electrodes, by deeplystudying different image reconstruction algorithms, the study that uses imagefusion to study the aging algorithm of electrical capacitance tomography systemis launched in this paper. The main works of this paper is as follows:Firstly, based on the research background, the significance of the researchtopic is discussed, the research status and the future trends of electricalcapacitance tomography are also studied in this paper.Secondly, based on the study of the components of ECT system, amathematical model of ECT system is established, it’s working principle and themathematical characteristics of ECT inverse problems are also studied. Based onthis, the study on different ECT image reconstruction algorithms is expanded,and some improvement works are used on the simultaneous iterative algorithm,and the simulation and analysis are also used on these algorithms.Again, studying on the spatial domain and transform domain, severaldifferent types of image fusion method are comparatively studied in this paper.Combing with the characteristics of the ECT system, the wavelet analysis isselected to study the ECT image fusion method. Finally, different fusion rules for high-frequency part and low-frequencypart are designed for the fusion of the each source image got from differentreconstruction algorithms, then the simulation experiment and the analyzes aremade for the images and the fusion rules. The rule based on edge detection isused in the high-frequency part. The weighted fusion rule based on algorithmanalysis is used in the low-frequency part. In this way, the image informationfrom different source image algorithms can be fully exploited, and the error ofthe final imaging results produced from source images also can be reduced inmaximum possible.