An Improved Newton-Raphson ert Image Reconstruction Method Based on Model Refinement

Abstract

The invention discloses a method for improving Newton-Raphson EPT image rebuilding quality based on model refining. According to the method, to solve the problem of unevenness of sensitivity distribution in an electrical resistance tomography technique, a finite element model b capable of effectively improving calculation precision of a direct problem is refined by the method that a node is added to the center position of each triangular finite element, and in the algorithm rebuilding process, the principle that when the direct problem is calculated, the finite element model b before refining is adopted, and when resistivity distribution is revised, a finite element model c after refining and a corresponding sensitivity matrix are adopted is followed. The method makes full use of advantages of the finite element b after optimization and advantages of the finite element c after the refining, on the basis that the calculation precision of the direct problem is effectively improved, evenness of the sensitivity distribution is improved, meanwhile morbidity performance of a Hessian matrix is improved, and on the premise that real-time performance of algorithm is not influenced, the image rebuilding quality is effectively improved.

Description

technical field [0001] The invention relates to electrical resistance tomography image reconstruction, in particular to an improved Newton-Raphson ERT image reconstruction method based on model refinement, belonging to the technical field of electrical tomography. Background technique [0002] Electrical tomography is an important branch of process tomography, including electrical resistance tomography (Electrical Resistance Tomography, ERT), electrical capacitance tomography (Electrical Capacitance Tomography, ECT), electrical impedance tomography (Electrical Impedance Tomography, EIT) and electromagnetic tomography (Electromagnetic Tomography, EMT), this technology has the advantages of fast response, non-invasive, low price, etc., and has broad application prospects in the fields of medicine and industrial measurement. [0003] In electrical tomography, the topology of the finite element model has an important impact on the calculation accuracy and image reconstruction qu...

AI Technical Summary

Problems solved by technology

Although the above refinement method can improve the accuracy, there are disadvantages at the same time: the element subdivision needs to impose local constraints at the suspended nodes, the calculation of the mesh reconstruction is very large, and the transformation from one mesh to another There are also some problems in grid division and transmission of data. r-refinement usually cannot achieve ideal accuracy, and is still in the theoretical research stage, while p-refinement cannot predict the order of polynomials in advance, and generally requires multiple recalculations , thus resulting in more computational cost

Images