| 引用本文: | 李家强， 陈焱博， 徐才秀， 陈金立， 刘然， 朱艳萍. 基于FGG NUFFT的穿墙成像雷达快速BP算法[J]. 雷达科学与技术, 2021, 19(1): 79-85.[点击复制] |
| LI Jiaqiang, CHEN Yanbo, XU Caixiu, CHEN Jinli, LIU Ran, ZHU Yanping. Fast BP Algorithm Based on FGG NUFFT for Through Wall Imaging Radar[J]. Radar Science and Technology, 2021, 19(1): 79-85.[点击复制] |
|
| 摘要: |
| 针对穿墙成像雷达后向投影(Back Projection，BP)算法存在计算复杂度较高、内存需求较大等问题，本文提出了一种基于快速高斯网格化的非均匀快速傅里叶变换(Fast Gaussian Gridding Nonuniform Fast Fourier Transform，FGG NUFFT)成像算法，该算法能够有效加速BP算法。对经过联合熵值法抑制墙体杂波后得到的目标回波数据，首先将BP算法中像素点幅值与高斯核函数反卷积消除高斯平滑的影响，然后对均匀数据进行快速傅里叶变换，最后对得到的数据进行卷积运算实现对数据均匀平滑输出。该方法预先划分网格并存储系数，避免了重复运算。通过对基于时域有限差分法(Finite Different-Time Domain，FDTD)的仿真软件GprMax2D/3D所获得的穿墙雷达数据进行处理，仿真实验证明该方法在保证成像质量的情况下，有效降低计算复杂度与内存需求。 |
| 关键词: 穿墙雷达 杂波抑制 BP算法 非均匀快速傅里叶变换 快速高斯网格化 |
| DOI：10.3969/j.issn.1672-2337.2021.01.013 |
| 分类号:TN957.51 |
| 基金项目:江苏高校品牌专业建设工程二期项目（电子信息工程）资助; 国家自然科学基金（No.61801231）; 江苏省自然科学基金（No.BK20191399） |
|
| Fast BP Algorithm Based on FGG NUFFT for Through Wall Imaging Radar |
|
LI Jiaqiang, CHEN Yanbo, XU Caixiu, CHEN Jinli, LIU Ran, ZHU Yanping
|
|
1.Collaborative Innovation Center on Forecast and Evaluation of Meteorological Disasters, Nanjing University of Information Science and Technology, Nanjing 210044,China;2.School of Electronics and Information Engineering, Nanjing University of Information Science and Technology, Nanjing 210044, China;3.Jiangsu Beidou Satellite Navigation Testing Center Corporation, Ltd, Nanjing 210032,China
|
| Abstract: |
| Aim to the problems of high computational complexity and large memory requirement for the back projection (BP) algorithm of through-wall imaging radar, a fast Gaussian gridding nonuniform fast Fourier transform is proposed to accelerate the calculation process of BP algorithm. Using the combination entropy method to suppress clutter from the wall, the influence of Gaussian smoothing is firstly eliminated by deconvolution between the BP algorithm of pixel amplitude and Gaussian kernel function. Then fast Fourier transform is applied to the uniform data. Finally the smooth output data are obtained by convolution operation. The proposed method divides the grid and stores the coefficients in advance to avoid repeated operation. Through processing the through-wall-radar data obtained by GprMax2D/3D, a simulation software based on FDTD, the simulation experiment proves that this method can effectively reduce the computational complexity and memory requirement under the condition of ensuring the imaging quality. |
| Key words: through-wall imaging radar clutter suppression BP algorithm nonuniform fast Fourier transform fast Gauss gridding |
