| 摘要: |
| 为了提高ISAR成像运算效率，本文提出一种自适应权重GPSR（AW GPSR）算法，并用于ISAR成像。该算法为加快参数重构的收敛速度，给ISAR图像中每个散射点都赋予正则化权重系数。对于幅度较小的散射系数，赋予较大的权重，使其快速减小为0；对于幅度较大的散射系数，给予较小的权重，使其在迭代中保持不变。在梯度下降迭代过程中，不断更新散射系数和正则化权重。本文采用仿真信号和实测信号进行ISAR成像实验，实验结果表明，相比常规GPSR算法，AW GPSR算法达到收敛的速度更快，成像时间降低了约23%。在成像效果方面，AW GPSR算法与常规GPSR算法相当，但明显好于传统的R?D算法和ESPRIT算法。 |
| 关键词: ISAR图像 压缩感知 GPSR算法 自适应权重 |
| DOI：DOI:10.3969/j.issn.1672-2337.2023.01.006 |
| 分类号:TN957.51 |
| 基金项目:国家自然科学基金（No.61890544，91748106） |
|
| ISAR Imaging Based on Adaptive Weight GPSR Algorithm |
|
SHI Runjia, HUANG Yifei, JIANG Zhongjin
|
|
State Key Laboratory of Millimeter Waves, Southeast University, Nanjing 210096, China
|
| Abstract: |
| In order to improve the efficiency of ISAR imaging, an adaptive weighted GPSR （AW GPSR） algorithm is proposed and applied to ISAR imaging. In order to accelerate the convergence speed of parameter reconstruction, each scattering point in ISAR image is given a regularization weight coefficient. For the scattering coefficient with small amplitude, a larger weight is given to make it quickly reduce to 0. For the scattering coefficient with large amplitude, a small weight is given to keep it unchanged in the next iterations. In the gradient descent iteration process, the scattering coefficient and regularization weight are continuously updated. In this paper, ISAR imaging experiments are carried out with simulated signals and measured signals. The experimental results show that AW GPSR algorithm converges faster and the imaging time is reduced by about 23% compared with the conventional GPSR algorithms. As to the imaging effect, AW GPSR algorithm is equivalent to the conventional GPSR algorithms, but obviously better than the traditional R?D algorithm and the ESPRIT algorithms. |
| Key words: ISAR image compressed sensing GPSR algorithm adaptive weight |
