سال انتشار: ۱۳۸۳

محل انتشار: سومین کنفرانس ماشین بینایی و پردازش تصویر

تعداد صفحات: ۵

نویسنده(ها):

M. Soleimani – The University of Manchester
A. Movafeghi – Sharif University of Technology

چکیده:

Electrical capacitance tomography (ECT) seeks to image the electrical permittivity distribution. ECT has potential applications in monitoring of twophase flows. The image reconstruction in ECT is an inverse medium problem and it is ill posed and nonlinear. Two-phase material reconstruction is a shape identification problem. Current states of the art techniques are pixel based image reconstruction methods, which are not effectively formulating the twophase property. In this paper we present a new interface based shape identification program using level set formulation. The level set method is a powerful technique in tracking interface propagations and has many applications including image processing. In this paper we study application of the level set method for a shape identification problem in ECT. The inverse boundary value problem of the low sensitive capacitance tomography imaging can be solved efficiently using level set method. Shape reconstruction method and especially formulation based on level set function can provide enough information to identify the object to be imaged. The technique reconstructs the interface between two phases. Main contributions of this paper are to introduce level set method to the inverse ECT problem and also reconstruction of the permittivity shape using experimental data.