Multiple arcs based image extrapolation method for millimeter wave UWB radar

Abstract

Range points migration (RPM) method is a successful 3-D imaging algorithm with higher accuracy and less computational cost. Due to the limitation of radar aperture size, usually, the boundary points estimated by RPM method only cover a small portion of target surface. Therefore image extrapolation method is needed for improving the image result of RPM method. The conventional image extrapolation method is established based on ellipsoid fitting. However, such a high freedom optimization problem leads to extremely high computational cost. In order to accelerate the computation, we propose a novel image extrapolation method for based on multiple arcs. In the proposed method, instead of ellipsoids, a group of arcs are estimated to describe the targets. In this way, original 3-D problem is degenerated to a 2-D one, which contributes a significant reduction of computational cost. The numerical simulation results show that the proposed method can realize image extrapolation in success with significantly higher accuracy and computational speed.