Research On ECT Image Reconstruction Algorithm Based On Compressed Sensing And Level Set

Process tomography technology originated in the 80 s of last century.After years of research and development,various forms of process tomography technology have been published.The success of these different mechanisms of tomography imaging system inspires the continuous research of researchers.Researchers from different disciplines apply these results to the study of two-phases(multi-phases)problems.Electrical capacitance tomography(ECT)is a process tomography technology that perceiving the distribution of media in pipes by capacitance plates.Electrical capacitance tomography(ECT)technology has become one of the mainstream methods for pipeline inspection because of its advantages such as simple structure,non intrusive and convenient measurement.At present,the image reconstruction algorithm of electrical capacitance tomography has not perfect yet,and the edge of the image is not clear enough.Therefore,it is of great practical significance to research a reasonable reconstruction algorithm of electrical capacitance tomography.In this paper,a large number of relevant literature are referred.The research status of the ECT image reconstruction algorithm at home and abroad is summarized.On the basis of the study of the related theory of capacitance tomography system,compressed sensing and level set,combining the characteristics of compressed sensing and level set method,a kind of sparsity adaptive compressed sensing and CV level set optimization ECT image construction algorithm is proposed.The following work is mainly done:1.Aiming at improving the unsatisfactory imaging effect of the existing ECT image reconstruction algorithm,an improved sparsity adaptive compressed sensing capacitance tomography algorithm is proposed.Firstly,the basic theory of compressed sensing and ECT image reconstruction is expressed,and the inverse problem solving model of capacitance tomography is derived.Secondly,the rationality of LBP algorithm image sparsity as the initial sparsity iteration value is verified,and a new sparsity adaptive compressed sensing algorithm is proposed.TheECT image's sparsity is selected and the image is reconstructed successfully.Finally,the experimental simulation is carried out,and the correlation coefficient and image error have been significantly improved.The feasibility of the algorithm is indicated.Compared with the existing compressed sensing method,the proposed method can accurately and automatically determine the sparsity of signals and has better signal recovery effect.2.Aiming at solving the problem of unclear image edge of ECT image reconstruction algorithm,a ECT image edge optimization method based on improved CV level set method is proposed.The method extracts the strongest inductive effect of ECT image based on the ability function,and achieves the effect of boundary optimization.Firstly,the basic idea of the level set method and the traditional CV level set model are briefly described.Secondly,the adaptability of the traditional CV level set model is improved based on the ECT image features.Finally,simulation experiments are carried out,and the correlation coefficient and image error are both improved,which verified the feasibility of the method.3.In order to further improve the reconstruction results of electrical capacitance tomography algorithm and make the edge of the image clearer,the improved adaptive compression perception algorithm and the CV level set edge optimization method are combined to reconstruct the image.Firstly,the initial imaging results are obtained by using the sparse adaptive compressed sensing algorithm.Secondly,the improved CV level set method is used to obtain the edge of the best reconstructed image and the image is reduced to the gray level,and the final reconstructed image results are obtained.Finally,the experimental simulation is carried out.The results show that the combined optimization algorithm's reconstruct the images are closer to the real distribution of the flow pattern than the existing image reconstruction algorithm,and two indicators which are the correlation coefficient and the image error are all improved.