Near-field cross section imaging based on 2-D NUFFT for millimeter wave
Near-field cross section imaging based on 2-D NUFFT for millimeter wave
复制标题
DOI：
10.1109/igarss.2016.7730320
复制
发表时间：
2016-07
期刊：
影响因子：
--
通讯作者：
Yingzhi Kan;Yongfeng Zhu;L. Tang;Hongzhong Zhao
中科院分区：
文献类型：
--
作者：
Yingzhi Kan;Yongfeng Zhu;L. Tang;Hongzhong Zhao
近场截面成像是近场雷达成像领域的一个重要课题。传统的截面成像算法是插值后的FFT算法。它是直观和相对快速的。然而，插值精度显着影响图像质量，特别是在宽带毫米波的情况下，当数据是高度弯曲的。为了解决这个问题，我们提出了一种基于二维非均匀快速傅立叶变换（2-D NUFFT）的算法。它将插值和FFT作为一个完整的过程，直接利用线阵上采集的数据。对相移后的二维空间谱数据进行二维NUFFT，得到高分辨率的近场图像。实验证明了该方法的有效性，与传统方法相比，该方法可以获得更精确的近场图像，且旁瓣更低。
Near-field cross section imaging is an important issue in the area of near filed radar imaging. A conventional algorithm for cross section imaging is FFT after interpolation. It is intuitionistic and relatively fast. However the interpolation accuracy significantly affects the image quality especially when the data is highly curving in the case of wideband millimeter wave. To solve this problem, we propose an algorithm based on two-dimensional nonuniform fast Fourier transform (2-D NUFFT). It considers interpolation and FFT as a whole process and directly exploits the data collected over a linear array. The 2-D NUFFT is performed over the phase shifted 2-D spatial spectrum data to obtain the near-field images with high resolution. Experiments certify the effectiveness of this method, which can achieve more accurate near-field images with lower sidelobes than the conventional method.