基于卷积神经网络的快速SAR自聚焦方法

doi:10.11947/j.AGCS.2024.20230281

测绘学报 ›› 2024, Vol. 53 ›› Issue (4): 610-619.doi: 10.11947/j.AGCS.2024.20230281

基于卷积神经网络的快速SAR自聚焦方法

刘志¹(), 杨淑媛¹(), 于子凡¹, 冯志玺¹, 高全伟¹, 王敏²

^1.西安电子科技大学人工智能学院智能感知与图像理解教育部重点实验室，陕西　西安　710071
^2.西安电子科技大学电子工程学院雷达信号处理全国重点实验室，陕西　西安　710071

收稿日期:2023-07-24 修回日期:2024-03-13 发布日期:2024-05-13
通讯作者: 杨淑媛 E-mail:zhiliu@stu.xidian.edu.cn;syyang@xidian.edu.cn
作者简介:刘志（1989—），男，博士，高级算法工程师，研究方向为深度学习和雷达信号处理。E-mail：zhiliu@stu.xidian.edu.cn
基金资助:
国家自然科学基金(62171357);陕西省科技厅自然科学基础研究计划面上项目(2023-JC-YB-524);陕西省教育科学“十三五”规划(SGH18H350)

Fast SAR autofocus based on convolutional neural networks

Zhi LIU¹(), Shuyuan YANG¹(), Zifan YU¹, Zhixi FENG¹, Quanwei GAO¹, Min WANG²

^1.Key Laboratory of Intelligent Perception and Image Understanding of Ministry of Education, School of Artificial Intelligence, Xidian University, Xi'an 710071, China
^2.National Key Laboratory of Radar Signal Processing, School of Electronic Engineering, Xidian University, Xi'an 710071, China

Received:2023-07-24 Revised:2024-03-13 Published:2024-05-13
Contact: Shuyuan YANG E-mail:zhiliu@stu.xidian.edu.cn;syyang@xidian.edu.cn
About author:LIU Zhi (1989—), male, PhD, senior algorithm engineer, majors in deep learning and radar signal processing. E-mail: zhiliu@stu.xidian.edu.cn
Supported by:
The National Natural Science Foundation of China(62171357);The Natural Science Basic Research Plan in Shaanxi Province of China(2023-JC-YB-524);The Education Scientific Program of the 13th Five-year Plan in Shaanxi Province of China(SGH18H350)

摘要/Abstract

摘要：

自聚焦是SAR高分辨率成像的关键技术。然而，传统的SAR自聚焦方法均需要迭代多次，实时性差，不适合在轨处理。本文提出了一种基于卷积神经网络的在轨快速SAR自聚焦方法（CNN-AF），该方法采用卷积神经网络来学习失焦图像到聚焦图像的映射，主要用于校正方位向的相位误差，由于在测试阶段该方法无须迭代和调整参数，因此该方法实时性好，更加适用于在轨处理。在真实SAR数据上的试验结果表明，本文方法具有较高的聚焦质量和聚焦速度。

关键词: 卷积神经网络, SAR, 相位误差, 自聚焦

Abstract:

Autofocus is a key technology for high-resolution synthetic aperture radar imaging. However, traditional SAR autofocus methods require too many iterations, have low computational efficiency, and are unsuitable for on-orbit processing. This paper proposes a fast SAR autofocus method based on convolutional neural networks. This method utilizes CNNs to learn the mapping from defocused images to focused images, mainly designed to correct the azimuth phase errors. It has a real-time performance and is more suitable for on-orbit processing since it does not need to iterate or adjust parameters in the testing phase. Experimental results on real SAR data show that our proposed method has the highest focusing quality and speed.

Key words: convolutional neural networks, SAR, phase error, autofocus

中图分类号:

P237

刘志, 杨淑媛, 于子凡, 冯志玺, 高全伟, 王敏. 基于卷积神经网络的快速SAR自聚焦方法[J]. 测绘学报, 2024, 53(4): 610-619.

Zhi LIU, Shuyuan YANG, Zifan YU, Zhixi FENG, Quanwei GAO, Min WANG. Fast SAR autofocus based on convolutional neural networks[J]. Acta Geodaetica et Cartographica Sinica, 2024, 53(4): 610-619.

图/表 9

图1

图2

图3

表1

图4

图5

图6

图7

表2

参考文献 31

[1]	李涛, 唐新明, 高小明, 等. SAR卫星业务化地形测绘能力分析与展望[J]. 测绘学报, 2021, 50(7):891-904. DOI: 10.11947/j.AGCS.2021.20200199.
	LI Tao, TANG Xinming, GAO Xiaoming, et al. Analysis and outlook of the operational topographic surveying and mapping capability of the SAR satellites[J]. Acta Geodaetica et Cartographica Sinica, 2021, 50(7):891-904. DOI: 10.11947/j.AGCS.2021.20200199.
[2]	刘银中. 机载SAR 实测数据多普勒参数估计与成像算法研究[D]. 哈尔滨: 哈尔滨工业大学, 2006.
	LIU Yinzhong. Research on Doppler parameter estimation and imaging algorithm of airborne SAR measured data [D]. Harbin: Harbin Industrial University, 2006.
[3]	王越. 合成孔径雷达自聚焦算法研究[D]. 南京: 南京航空航天大学, 2006.
	WANG Yue. Research on autofocus algorithm of synthetic aperture radar[D]. Nanjing: Nanjing University of Aeronautics and Astronautics, 2006.
[4]	KIM J W, KIM Y D, YEO T D, et.al. Fast Fourier-domain optimization using hybrid L₁-L_p-Norm for autofocus in airborne SAR imaging[J]. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(10):7934-7954.
[5]	IAN G C, FRANK H C W. Digital processing of synthetic aperture radar data: algorithms and implementation[M]. Norwood: Artech House, 2005: 1-436.
[6]	SAMCZYNSKI P, KULPA K S. Coherent mapdrift technique[J]. IEEE Transactions on Geoscience and Remote Sensing, 2009, 48(3):1505-1517.
[7]	CALLOWAY T, DONOHOE G. Subaperture autofocus for synthetic aperture radar[J]. IEEE Transactions on Aerospace and Electronic Systems, 1994, 30(2):617-621.
[8]	YANG Ruliang, LI Haiying, LI Shiqiang, et al. High-resolution microwave imaging[M]. Singapore: Springer, 2018: 1-569.
[9]	LIU Zhi, YANG Shuyuan, FENG Zhixi, et al. Fast SAR autofocus based on ensemble convolutional extreme learning machine[J]. Remote Sensing, 2021, 13(14):2683.
[10]	LI Xi, LIU Guosui, NI Jinlin. Autofocusing of ISAR images based on entropy minimization[J]. IEEE Transactions on Aerospace and Electronic Systems, 1999, 35(4):1240-1252.
[11]	CAI Jinjian, MARTORELLA M, CHANG Shaoqiang, et al. Efficient nonparametric ISAR autofocus algorithm based on contrast maximization and Newton's method[J]. IEEE Sensors Journal, 2020, 21(4):4474-4487.
[12]	MORRISON R L, DO M N, MUNSON D C. SAR image autofocus by sharpness optimization: a theoretical study[J]. IEEE Transactions on Image Processing, 2007, 16(9):2309-2321.
[13]	ZHANG Shuanghui, LIU Yongxiang, LI Xiang. Fast entropy minimization based autofocusing technique for ISAR imaging[J]. IEEE Transactions on Signal Processing, 2015, 63(13):3425-3434.
[14]	ZENG Tao, WANG Rui, LI Feng. SAR image autofocus utilizing minimum-entropy criterion[J]. IEEE Geoscience and Remote Sensing Letters, 2013, 10(6):1552-1556.
[15]	WAHL D E, EICHEL P H, GHIGLIA D C, et al. Phase gradient autofocus: a robust tool for high resolution SAR phase correction[J]. IEEE Transactions on Aerospace and electronic systems, 1994, 30(3):827-835.
[16]	卿吉明, 徐浩煜, 梁兴东, 等. 一种可用于实时成像的改进PGA算法[J]. 雷达学报, 2015, 4(5):600-607.
	QING Jiming, XU Haoyu, LIANG Xingdong, et al. An improved phase gradient autofocus algorithm used in real-time processing[J]. Journal of Radars, 2015, 4(5):600-607.
[17]	LEE H, JUNG C S, KIM K W. Feature preserving autofocus algorithm for phase error correction of SAR images[J]. Sensors, 2021, 21(7):2370.
[18]	郭华东, 吴文瑾, 张珂, 等. 新型SAR对地环境观测[J]. 测绘学报, 2022, 51(6):862-872. DOI: 10.11947/j.AGCS.2022.20220098.
	GUO Huadong, WU Wenjin, ZHANG Ke, et al. New generation SAR for Earth environment observation[J]. Acta Geodaetica et Cartographica Sinica, 2022, 51(6):862-872. DOI: 10.11947/j.AGCS.2022.20220098.
[19]	QIAO Guangkai, DAI Jingwei, WANG Kaizhi, et al. A finely focusing method of SAR using very deep neural network[C]//Proceedings of the 39th IEEE International Geoscience and Remote Sensing Symposium. Yokohama: IEEE, 2019: 2571-2574.
[20]	LIU Zhi, YANG Shuyuan, GAO Quanwei, et al. AFnet and PAFnet: fast and accurate SAR autofocus based on deep learning[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022(60):1-13.
[21]	HUO Jiawei, LI Min, LI Zhongyu, et al. Novel SAR autofocus method based on deep learning with GAN network[C]//Proceedings of 2021 CIE International Conference on Radar. Haikou: IEEE, 2021: 800-803.
[22]	WEI Pu. Deep learning algorithm for SAR autofocus[C]//Proceedings of 2021 General assembly and scientific symposium of the international union of radio science. Rome: IEEE, 2021: 1-4.
[23]	AHMED A E A, LI Zhenfang. Improved phase gradient autofocus algorithm based on segments of variable lengths and minimum-entropy phase correction[J]. IET Radar, Sonar & Navigation, 2015, 9(4):467-479.
[24]	LI Y, O'YOUNG S. Kalman filter disciplined phase gradient autofocus for stripmap SAR[J]. IEEE Transactions on Geoscience and Remote Sensing, 2020, 58(9):6298-6308.
[25]	温雪娇, 仇晓兰, 尤红建, 等. 高分辨率星载SAR起伏运动目标精细聚焦与参数估计方法[J]. 雷达学报, 2017, 6(2):213-220. DOI: 10.12000/JR17005.
	WEN Xuejiao, QIU Xiaolan, YOU Hongjian, et al. Focusing and parameter estimation of fluctuating targets in high resolution spacebone SAR[J]. Journal of Radars, 2017, 6(2):213-220. DOI: 10.12000/JR17005.
[26]	LECUN Y, BOTTOU L, BENGIO Y, et al. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11):2278-2324.
[27]	NAIR V, HINTON G E. Rectified linear units improve restricted boltzmann machines[C]//Proceedings of the 27th International Conference on Machine Learning. Haifa: IEEE, 2010: 807-814.
[28]	TAI Y, YANG J, LIU X. Image super-resolution via deep recursive residual network[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition. Hawaii: IEEE, 2017: 3147-3155.
[29]	BOTTOU L. Stochastic gradient descent tricks[C]//Proceedings of 2012 Neural Networks. Berlin: Springer, 2012.
[30]	KINGMA D P, BA J. Adam: a method for stochastic optimization[C]//Proceedings of the 3th International Conference on Learning Representations. San Diego: [s.n.], 2015.
[31]	LOSHCHILOV I, HUTTER F. Decoupled weight decay regularization[C]//Proceedings of the 7th International Conference on Learning Representations. Vancouver: [s.n.], 2019.

方法	熵	对比度	CPU耗时/s	GPU耗时/s
PGA-ML	9.7982	5.976 8	1 597.15	3 351.76
PGA-LUMV	9.796 5	5.983 6	1 617.23	3 555.88
MEA	9.759 3	6.254 9	4 511.39	938.35
N-MEA^[11]	9.761 3	6.248 9	1 469.83	260.06
VDNN-AF^[19]	9.637 9	7.164 4	3 656.93	156.69
CNN-AF	9.578 0	7.402 3	780.50	47.10

方法	熵	对比度	CPU耗时/s	GPU耗时/s
PGA-ML	9.878 7	4.805 6	566.43	1 112.43
PGA-LUMV	9.875 5	4.831 2	585.19	1 108.25
MEA（0.01）	9.896 0	4.598 1	3 491.33	217.56
MEA（0.1）	9.846 7	5.071 0	3 315.42	217.67
MEA（1）	9.844 8	4.598 1	3 272.45	217.31
MEA（10）	9.845 1	5.096 1	3 304.63	216.28
MEA（100）	10.077 3	3.637 1	3 251.51	214.41
N-MEA^[11]	9.836 0	5.212 5	616.77	38.94
AFnet^[20]	9.852 8	4.993 4	369.67	17.61
PAFnet^[20]	9.852 4	5.000 5	203.03	12.05
VDNN-AF^[19]	9.637 9	7.164 4	1 315.94	52.40
CNN-AF	9.578 0	7.402 3	186.31	17.07

基于卷积神经网络的快速SAR自聚焦方法

Fast SAR autofocus based on convolutional neural networks

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 9

参考文献 31

相关文章 15

编辑推荐

Metrics

本文评价

[1]	江利明, 邵益, 周志伟, 马培峰, 王腾. 智能化InSAR数据处理研究进展、挑战与展望[J]. 测绘学报, 2024, 53(6): 1037-1056.
[2]	龚循强, 汪宏宇, 鲁铁定, 游为. 高铁桥墩沉降的通用渐进分解长期预测网络模型[J]. 测绘学报, 2024, 53(6): 1113-1127.
[3]	周苍海, 田镇, 石震, 托坎哈衣那尔·. 时序InSAR与GNSS联合约束下的亚东-谷露断裂运动特征[J]. 测绘学报, 2024, 53(5): 933-945.
[4]	朱珺, 彭葳, 付海强, 欧蔓, 雷缮诚, 张石平. 观测信息不足条件下TanDEM-X InSAR的大范围林下地形反演[J]. 测绘学报, 2024, 53(5): 959-966.
[5]	高延东, 贾义琨, 李世金, 陈宇, 李怀展, 郑南山, 张书毕. InSAR相位解缠最大流/最小割权值改进算法[J]. 测绘学报, 2024, 53(4): 644-652.
[6]	黄启灏, 靳国旺, 熊新, 王丽美, 李佳豪. 通道剪枝与知识蒸馏相结合的轻量化SAR目标检测[J]. 测绘学报, 2024, 53(4): 712-723.
[7]	何毅, 杨旺, 朱庆. 基于R2AU-Net的InSAR相位解缠方法[J]. 测绘学报, 2024, 53(3): 435-449.
[8]	项德良, 丁怀跃, 管冬冬, 程建达, 孙晓坤. 结合极化白化滤波和SimSD-CapsuleNet的PolSAR图像配准[J]. 测绘学报, 2024, 53(3): 450-462.
[9]	王媛, 徐华平, 李春升, 曾国兵, 刘爱芳, 葛仕奇. 星载分布式SAR双频乒乓模式测高精度分析[J]. 测绘学报, 2024, 53(3): 463-472.
[10]	岳佳伟, 黄其欢, 刘辉, 马张烽. 离散优化框架下的多基线InSAR相位解缠方法[J]. 测绘学报, 2024, 53(3): 473-481.
[11]	高贤文, 金涛勇, 李建成. 兼顾内陆水体水位时空变化特性的合成孔径雷达波形重跟踪算法[J]. 测绘学报, 2024, 53(2): 217-230.
[12]	翟振和, 孙中苗, 管斌, 马健, 李端. 利用扰动重力数据反演海底地形的高斯曲面函数估计方法[J]. 测绘学报, 2024, 53(2): 231-238.
[13]	姜三, 刘凯, 李清泉, 江万寿. 融合深度特征的无人机影像SfM重建[J]. 测绘学报, 2024, 53(2): 321-331.
[14]	廖钊宏, 张依晨, 杨飚, 林明春, 孙文博, 高智. 基于Swin Transformer-CNN的单目遥感影像高程估计方法及其在公路建设场景中的应用[J]. 测绘学报, 2024, 53(2): 344-352.
[15]	孙一帆, 刘冰, 余旭初, 谭熊, 余岸竹. 图像级高光谱影像高分辨率特征网络分类方法[J]. 测绘学报, 2024, 53(1): 50-64.