Real time reconstruction of 3D cylindrical near field radar images using a single instruction multiple data interpolation approach

摘要

The present invention uses a Single Instruction Multiple Data (SIMD) architecture to form real time 3D radar images recorded in cylindrical near field scenarios using a wavefront reconstruction approach. A novel interpolation approach is executed in parallel, significantly reducing the reconstruction time without compromising the spatial accuracy and signal to noise ratios of the resulting images. Since each point in the problem space can be processed independently, the proposed technique was implemented using an approach on a General Purpose Graphics Processing Unit (GPGPU) to take advantage of the high performance computing capabilities of this platform.

主权项

1. A method of reconstructing collected data acquired in a three dimensional cylindrical radar scan geometry in which targets are considered to be at near field distances and where the scan planes are defined along a z axis, scan locations are arranged in a circular pattern with a radius R at each plane, an irradiating antenna is facing towards a center of each scan plane, and a waveform f(t) with a prescribed bandwidth is sequentially radiated from each scan location, the method comprising:
expressing a received signal from each scan location (R, θ, z) as:s⁡(t,θ,z)=∑q=1T⁢σq⁢f⁡(t-2⁢⁢Dq⁡(θ,z)v)where νis the propagation speed, σq and (rp, φq, zp) are the reflectivity and location of the qth target and
Dq(θ, z)=√{square root over (R2+rq2+(zq−z)2−2Rrq cos(φq−θ))}{square root over (R2+rq2+(zq−z)2−2Rrq cos(φq−θ))}; calculating a Fourier transform along the t, z and θ directions of said received signal; applying a filter such that a resulting compensated dataset U(ω, ∈, kz) is expressed as:U⁡(ω,ɛ,kz)=∑q=1T⁢4⁢⁢σq·ξ⁡(ω,ɛ,kz)·exp⁡(-j⁡((4⁢⁢k2-kz2)⁢rq2-ɛ2)+ɛ·sin-1⁡(ɛ/(4⁢⁢k2-kz2)⁢rq2)+ɛ⁢⁢ϕq+kz⁢zq)where ∈ and kz are the spatial frequency counterparts of θ and z respectively and ξ(ω, ∈, kz) is the spectrum amplitude component in the (ω, ∈, kz) frequency space;
applying a mapping function to the resulting compensated dataset to produce a sampled frequency space denoted as I(kux, kuy, kz); interpolating values of I(kux, kuy, kz) at points defined in (kx, ky, kz); and calculating a three dimensional inverse Fourier transform of I(kx, ky, kz) to yield a three dimensional model i(x, y, z).