M-FCN based sea-surface weak target detection

doi:10.23919/JSEE.2021.000095

Journal of Systems Engineering and Electronics ›› 2021, Vol. 32 ›› Issue (5): 1111-1118.doi: 10.23919/JSEE.2021.000095

• ELECTRONICS TECHNOLOGY • Previous Articles Next Articles

M-FCN based sea-surface weak target detection

Meiyan PAN^1,^2,*(), Jun SUN^1,²(), Yuhao YANG^1,²(), Dasheng LI^1,²(), Junpeng YU^1,²()

¹ The 14th Research Institute of China Electronics Technology Group Corporation, Nanjing 210039, China
² Key Laboratory of IntelliSense Technology, China Electronics Technology Group Corporation, Nanjing 210039, China

Received:2020-09-07 Online:2021-10-18 Published:2021-11-04
Contact: Meiyan PAN E-mail:meiyan_pan@163.com;sunjun@ustc.edu;yyhao@mail.ustc.edu.cn;lds412@163.com;yjp603@163.com
About author:|PAN Meiyan was born in 1993. She is a master and an assistant engineer in Nanjing Research Institute of Electronics Technology. Her research interests are marine target detection, sea clutter suppression, and deep learning. E-mail: meiyan_pan@163.com||SUN Jun was born in 1974. He is a Ph.D. and the minister of the Key Laboratory of IntelliSense Technology, China Electronics Technology Group Corporation (CETC). His research interests are radar signal processing, target recognition, and pattern recognition. E-mail: sunjun@ustc.edu||YANG Yuhao was born in 1983. He is a Ph.D. and a senior engineer in Nanjing Research Institute of Electronics Technology. His research interests are radar signal processing, radar imaging, and target recognition. E-mail: yyhao@mail.ustc.edu.cn||LI Dasheng was born in 1982. He is a Ph.D. and a researcher-level senior engineer in Nanjing Research Institute of Electronics Technology. His research interests are Terahertz radar system, radar target characteristics, and radar system simulation. E-mail: lds412@163.com||YU Junpeng was born in 1984. He is a master and a senior engineer in Nanjing Research Institute of Electronics Technology. His research interests are radar signal processing and SAR imaging technology. E-mail: yjp603@163.com
Supported by:
This was work supported by the National Natural Science Foundation of China (U19B2031).

Abstract

Abstract:

This paper focuses on the sea-surface weak target detection based on memory-fully convolutional network (M-FCN) in strong sea clutter. Firstly, the constant false alarm rate (CFAR) detection method utilizes a low threshold with high probability of false alarm to detect sea-surface weak targets after non-coherent integration. Reducing the detection threshold can generate a large number of false alarms while increasing the detection rate, and how to suppress a large number of false alarms is the key to improve the performance of weak target detection. Then, the detection result of the low threshold is operated to construct the target matrix suitable for the size of fully convolutional networks and the convolution operator form. Finally, the M-FCN architecture is designed to learn the different accumulation characteristics of the target and the sea clutter between different frames. For improving the detection performance, the historical multi-frame information is memorized by the network, and the end-to-end structure is established to detect sea-surface weak target automatically. Experimental results on measured data demonstrate that the M-FCN method outperforms the traditional track before detection (TBD) method and reduces false alarm tracks by 35.1%, which greatly improves the track quality.

Key words: sea-surface weak target detection, memory-fully convolutional network (M-FCN), multi-frame information, end-to-end

Meiyan PAN, Jun SUN, Yuhao YANG, Dasheng LI, Junpeng YU. M-FCN based sea-surface weak target detection[J]. Journal of Systems Engineering and Electronics, 2021, 32(5): 1111-1118.

Figures/Tables 11

Fig 1

Fig 2

Table 1

Fig 3

Fig 4

Fig 5

Fig 6

Fig 7

Fig 8

Fig 9

Fig 10

References 27

1	DING L F, GENG G F. Radar principle. Xi’an: Xidian University Press, 2002. (in Chinese)
2	XUE J, XU S W, LIU J, et al Model for non-Gaussian sea clutter amplitudes using generalized inverse Gaussian texture. IEEE Geoscience and Remote Sensing Letters, 2019, 16 (6): 892- 896. doi: 10.1109/LGRS.2018.2886782
3	HE Y, HUANG Y, GUAN J, et al. An overview on radar target detection in sea clutter. Modern Radar, 2014, 36(12): 6−14. (in Chinese)
4	LI Q Q. Radar target detection and performance evaluation in sea clutter. Xi’an: Xidian University, 2014. (in Chinese)
5	XUE J, XU S W, SHUI P L Near-optimum coherent CFAR detection of radar targets in compound-Gaussian clutter with inverse Gaussian texture. Signal Processing, 2020, 166, 107236. doi: 10.1016/j.sigpro.2019.07.029
6	SAMUEL J D, MARK G R, BRAIN C. A comparison of detection performance for several track-before-detect algorithms. Eurasip Journal on Advances in Signal Processing, 2008. DOI: 10.1155/2008/428036.
7	XUE J, XU S W, SHUI P L Improved track-before-detect method for detecting range-spread targets in generalized Pareto clutter. Science China-Information Sciences, 2019, 62, 40304. doi: 10.1007/s11432-018-9744-2
8	DING H, DONG Y L, LIU N B, et al Overview and prospects of research on sea clutter property cognition. Journal of Radars, 2016, 5 (5): 499- 516. doi: 10.12000/JR16069
9	XU S W, SHI X Y, XUE J, et al Adaptive subspace detection of range-spread target in compound Gaussian clutter with inverse Gaussian texture. Digital Signal Processing, 2018, 81, 79- 89. doi: 10.1016/j.dsp.2018.07.002
10	WU S J, MEI X C Radar signal processing and data processing technology. Beijing: Publishing House of Electronic Industry, 2007.
11	LONG J, SHELHAMER E, DARRELL T Fully convolutional networks for semantic segmentation. Proc. of the 28th IEEE Conference on Computer Vision and Pattern Recognition, 2015, 3431- 3440.
12	KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks, in Proceedings of the 25th Int Conf on Neural Information Processing Systems. Cambridge, MA: MIT Press, 2012: 1097-1105.
13	REN S Q, HE K M, GIRSHICK R, et al Faster RCNN: towards real-time object detection with region proposal networks. IEEE Trans. on Pattern Analysis and Machine Intelligence, 2015, 39 (6): 1137- 1149.
14	NOH H, HONG S, HAN B Learning deconvolution network for semantic segmentation. Proc. of the 15th IEEE International Conference on Computer Vision, 2015, 1520- 1528.
15	SZARVAS M, YOSHIZAWA A, YAMAMOTO M Multi-view face detection using deep convolutional neural networks. Proc. of the 5th ACM on International Conference on Multimedia Retrieval, 2015, 643- 650.
16	HUANG L C, YANG Y, DENG Y F, et al. Dense Box: unifying landmark localization with end to end object detection. arXiv, 2016: 1509.04874.
17	DAI J F, HE K M, SUN J. Instance-aware semantic segmentation via multi-task network cascades, in Proceedings of the 29th IEEE Conf on Computer Vision and Pattern Recognition. Piscataway, NJ: IEEE, 2016: 3150-3158.
18	WANG H Q, JIANG B, LI X, et al The method of target detection based on ameliorated SVM in the presence of sea clutter. Signal Processing, 2007, 23 (4): 598- 602.
19	DURKIN J, CAI J F, CAI Z X Decision tree technique and its current research. Control Engineering of China, 2005, 12 (1): 15- 18.
20	SONG C Y. Hyperspectral remote sensing image processing system based on fully convolutional network. Nanjing: Nanjing University of Science and Technology, 2018. (in Chinese)
21	MATTHEW D Z, DILIP K, GRAHAM W T, et al. Deconvolutional networks. Proc. of the IEEE Conference on Compute Vision Pattern Recognition, 2010: 2528−2535.
22	ROBBINS H, MONRO S. A stochastic approximation method. New York: Springer, 1951.
23	DUCHI J, HAZAN E, SINGER Y Adaptive sub-gradient methods for online learning and stochastic optimization. Journal of Machine Learning Research, 2011, 12 (7): 2121- 2159.
24	ZEILER M D. ADADELTA: an adaptive learning rate method. http://arXiv.org/pdf/1212.5701.pdf.
25	HINTON G Lecture 6.5-rmsprop: divide the gradient by a running average of its recent magnitude. Neural Networks for Machine Learning, 2012, 4 (2): 26- 31.
26	KINGMA D P, BA J A. A method for stochastic optimization. http://arXiv.org/pdf/1412.6980.pdf.
27	RUDER S. An overview of gradient descent optimization algorithms. http://arXiv.org/pdf1412.6980/.pdf.

Hardware	Attribute	Software	Attribute
Operating system	Windows 7	Programming language	Python 3.5
RAM	128 GB	Integrated development environment	PyCharm
Graphics card	NAVID TITAN V	Deep learning framework	Keras
Video memory	12 GB	Programming software	Matlab2018

M-FCN based sea-surface weak target detection

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

Share this article

Figures/Tables 11

References 27

Related Articles 1

Recommended Articles

Metrics

Comments