MICROWAVE IMAGING OF DIELECTRIC CYLINDERS USING LEVEL SET METHOD AND CONJUGATE GRADIENT ALGORITHM

In this paper, a level set method for shape reconstruction problems is considered. By measuring the scattered fleld, we tried to retrieve the localisation and permittivity of buried objects. The forward problem is solved by the method of moments. For solving the inverse problem, we adopt an evolution approach. Therefore, we introduce a level set technique witch is ∞exible in handling complex shape changes. A conjugate gradient-based method is used in order to deflne iterative updates for the level set functions with the goal to minimize a given least squares data misflt functional. In particular, the proposed method is capable of creating new holes inside the design domain, which makes the flnal design independent of initial guess and reduces the probability of converging to a local minimum. Experimental results demonstrate the feasibility and efiectiveness of the proposed technique.