Study On Image Reconstruction Algorithms Of Region Of Interest For Electrical Capacitance Tomography

With the characteristics of low-cost inputs, wide application fields, simplestructure, non-intrusive testing method and good safety performance, ECT(Electrical Capacitance Tomography) has been the focus of the study on the flowimaging. But some disadvantages of the ECT has hindered the continuingdevelopment of the image reconstruction. Such as nonlinear features of ECT,limited measurement data of capacitance, fewer data for image reconstruction, noanalytical solution forinverseproblem, Instability and morbid of the whole systemand so on. The image reconstruction algorithm is a key difficulty affecting thefurther development and practicalization of ECT, so, the study on searching areasonable algorithm appears extremely important and urgent.This paper focuses on the reconstruction algorithms for electrical capacitancetomography about region of interest, the main works can be summarized asfollowing:1. Taking the image reconstruction of oil-water two-phase flow system inECT system as research object, the paper introduces the system structure,working principle, characteristics and defectsof ECT system. Then themathematical model and sensing fieldis analyzed to make ready for the followingresearch.2. In order to provide the theory basis and the technical support for imageformation in some regions of interest, the existingprinciple and algorithmimplementation of ECT system are analyzed and discussed.3. The theory and present application of compressed sensing (CS) areintroduced at fiirst. Then, based on the ART (Algebraic ReconstructionTechnique), a new image reconstruction algorithm combining CS with ART is proposed, which can obtain a higher accuracy imaging by combining the sparsityof image gradient with projectiondata and using total variation to reduce the normof image gradient. It’s a new approach to image reconstruction in ECT system.4. This thesis takes ECT system of12-electrodes as research object, an imagereconstruction experiment respectively adopting the original and improved imagereconstruction algorithm is simulated to compare the imaging result by two ways.The result proves that the image reconstruction algorithm proposed in this paperobtain a higher accuracy and has some practicability.