生成式中断航迹接续关联方法

doi:10.12305/j.issn.1001-506X.2022.05.15

摘要/Abstract

摘要：

传统中断航迹接续关联(track segment consecutive association, TSCA)方法基于假设的目标运动模型, 利用大量先验信息完成关联任务, 存在参数过多、计算复杂、推理时间长等缺点。为了解决以上问题, 提出一种基于注意力机制的生成式TSCA方法。首先设计航迹态势图生成模块, 将原始航迹数据转换为航迹态势图, 作为生成对抗网络的输入。针对航迹噪声影响大和航迹运动特征、中断特征难以提取的问题, 基于生成对抗网络和注意力机制, 设计航迹关联网络, 滤除航迹噪声并完成TSCA。仿真结果证明了所提网络的有效性, 在关联精度和关联速度两方面都超过现有算法。

关键词: 中断航迹接续关联, 生成对抗网络, 注意力机制, 航迹态势图

Abstract:

Traditional track segment consecutive association (TSCA) methods are based on the hypothesis target motion model and need to use a lot of prior information. It has many disadvantages such as too many parameters, complicated calculation and long reasoning time. In order to solve the above problems, a generative TSCA method based on the attention mechanism is proposed. Firstly, the module of generating the track situation map is designed, and the original track data is converted into the track situation map as the input of generative adversarial network. Aiming at the problems of track noise and the difficulty in extracting the features of track motion and track interruption, based on the generative adversarial network and the attention mechanism, the track association network is designed to filter out the noise of tracks and accomplish TSCA. The simulation results show that the proposed method is effective and exceeds the existing algorithms in both precision and speed.

Key words: track segment consecutive association (TSCA), generative adversarial network, attention mechanism, track situation map

中图分类号:

徐平亮, 崔亚奇, 熊伟, 熊振宇, 顾祥岐. 生成式中断航迹接续关联方法[J]. 系统工程与电子技术, 2022, 44(5): 1543-1552.

Pingliang XU, Yaqi CUI, Wei XIONG, Zhenyu XIONG, Xiangqi GU. Generative track segment consecutive association method[J]. Systems Engineering and Electronics, 2022, 44(5): 1543-1552.

图/表 18

图1

图2

图3

图4

图5

图6

图7

图8

表1

表2

图9

表3

图10

图11

图12

图13

表4

图14

参考文献 33

1	YEOM S W, KIRUBARAJAN T, BAR S Y. Improving track continuity using track segment association[C]//Proc. of the IEEE Aerospace Conference, 2003.
2	SUN J P, WANG N Y, ZHANG Z G. Track segment association of maneuvering target based on expectation maximization[C]//Proc. of the 11th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, 2018.
3	ZHANG S , BAR S Y . Track segment association for GMTI tracks of evasive move-stop-move maneuvering targets[J]. IEEE Trans.on Aerospace and Electronic Systems, 2011, 47 (3): 1899- 1914. doi: 10.1109/TAES.2011.5937272
4	DEMPSTER A P , LAIRD N M , RUBIN D B . Maximum likelihood from incomplete data via the EM algorithm[J]. Journal of the Royal Statistical Society: Series B (Methodological), 1977, 39 (1): 1- 22. doi: 10.1111/j.2517-6161.1977.tb01600.x
5	ZHU H Y , HAN S Y . Track-to-track association based on structural similarity in the presence of sensor biases[J]. Journal of Applied Mathematics, 2014, 294657.
6	杜渐, 夏学知. 面向航迹中断的模糊航迹关联算法[J]. 火力与指挥控制, 2013, 38 (6): 68- 71.
	DU J , XIA X Z . A fuzzy track association algorithm in track interrupt-oriented[J]. Fire Control & Command Control, 2013, 38 (6): 68- 71.
7	刘颢, 陈世友, 汪学东, 等. 一种自适应航迹关联算法[J]. 电子学报, 2013, 41 (12): 2416- 2421. doi: 10.3969/j.issn.0372-2112.2013.12.015
	LIU H , CHEN S Y , WANG X D , et al. An adaptive track correlation algorithm[J]. Chinese Journal of Electronics, 2013, 41 (12): 2416- 2421. doi: 10.3969/j.issn.0372-2112.2013.12.015
8	齐林, 王海鹏, 熊伟, 等. 基于先验信息的多假设模型中断航迹关联算法[J]. 系统工程与电子技术, 2015, 37 (4): 732- 739.
	QI L , WANG H P , XIONG W , et al. Track segment association algorithm based on multiple-hypothesis models with priori information[J]. Journal of Systems Engineering and Electronics, 2015, 37 (4): 732- 739.
9	DONG H, YU S M, WU C, et al. Semantic image synthesis via adversarial learning[C]//Proc. of the IEEE International Conference on Computer Vision, 2017: 5706-5714.
10	KANEKO T, HIRAMATSU K, KASHINO K. Generative attribute controller with conditional filtered generative adversarial networks[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017.
11	KARACAN L, AKATA Z, ERDEM A, et al. Learning to generate images of outdoor scenes from attributes and semantic layouts[EB/OL]. [2021-01-25]. https://arXiv.org/pdf/1612.00215.pdf.
12	LEDIG C, THEIS L, HUSZAR F, et al. Photo-realistic single image super-resolution using a generative adversarial network[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 4681-4690.
13	PATHAK D, KRAHENBUHL P, DONAHUE J, et al. Context encoders: feature learning by inpainting[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 2536-2544.
14	SANGKLOY P, LU J W, FANG C, et al. Scribbler: controlling deep image synthesis with sketch and color[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 5400-5409.
15	WANG X L, GUPTA A. Generative image modeling using style and structure adversarial networks[C]//Proc. of the European Conference on Computer Vision, 2016: 318-335.
16	ZHANG Z F, SONG Y, QI H R. Age progression/regression by conditional adversarial autoencoder[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 5810-5818.
17	GOODFELLOW I , POUGET A J , MIRZA M , et al. Generative adversarial nets[J]. Advances in Neural Information Processing Systems, 2014, 27, 2672- 2680.
18	MNIH V , HEESS N , GRAVES A . Recurrent models of visual attention[J]. Advances in Neural Information Processing Systems, 2014, 2, 2204- 2212.
19	WANG K F , GOU C , DUAN Y J , et al. Generative adversa-rial networks: the state of the art and beyond[J]. Acta Automatica Sinica, 2017, 43 (3): 321- 332.
20	VINCENT P, LAROCHELLE H, BENGIO Y, et al. Extracting and composing robust features with denoising autoencoders[C]//Proc. of the 25th International Conference on Machine Learning, 2008: 1096-1103.
21	ULYANOV D, VEDALDI A, LEMPITSKY V. Instance normalization: the missing ingredient for fast stylization[EB/OL]. [2021-01-25]. https//arXiv.org/abs/1607.08022v1.
22	GLOROT X, BORDES A, BENGIO Y. Deep sparse rectifier neural networks[C]//Proc. of the 14th International Confe-rence on Artificial Intelligence and Statistics, 2011: 315-323.
23	HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 770-778.
24	SIMONYAN K, ZISSERMAN A. Very deep convolutional networks for large-scale image recognition[EB/OL]. [2021-01-25]. https//arXiv.org/abs/1409.1556.
25	DUMOULIN V, VISIN F. A guide to convolution arithmetic for deep learning[EB/OL]. [2021-01-25]. https//arXiv.org/abs/1603.07285.
26	PENG X B, KANAZAWA A, TOYER S, et al. Variational discriminator bottleneck: improving imitation learning, inverserl, and gans by constraining information flow[EB/OL]. [2021-01-25]. https//arXiv.org/abs/1810.00821v4.
27	RATLIFF L J, BURDEN S A, SASTRY S S. Characterization and computation of local Nash equilibria in continuous games[C]//Proc. of the IEEE 51st Annual Allerton Conference on Communication, Control, and Computing, 2013: 917-924.
28	ZHAO H , GALLO O , FROSIO I , et al. Loss functions for image restoration with neural networks[J]. IEEE Trans.on Computational Imaging, 2016, 3 (1): 47- 57.
29	LI X R , JILKOV V P . Survey of maneuvering target tracking. Part I. dynamic models[J]. IEEE Trans.on Aerospace and Electronic Systems, 2003, 39 (4): 1333- 1364. doi: 10.1109/TAES.2003.1261132
30	WANG Z , BOVIK A C , SHEIKH H R , et al. Image quality assessment: from error visibility to structural similarity[J]. IEEE Trans.on Image Processing, 2004, 13 (4): 600- 612. doi: 10.1109/TIP.2003.819861
31	PASZKE A, GROSS S, MASSA F, et al. Pytorch: an imperative style, high-performance deep learning library[EB/OL]. [2021-01-25]. https//arXiv.org/abs/1912.01703v1.
32	YEOM S W , KIRUBARAJAN T , BAR S Y . Track segment association, fine-step IMM and initialization with Doppler for improved track performance[J]. IEEE Trans.on Aerospace and Electronic Systems, 2004, 40 (1): 293- 309. doi: 10.1109/TAES.2004.1292161
33	RAGHU J , SRIHARI P , THARMARASA R , et al. Comprehensive track segment association for improved track continuity[J]. IEEE Trans.on Aerospace and Electronic Systems, 2018, 54 (5): 2463- 2480. doi: 10.1109/TAES.2018.2820364

输出通道	AP	P@5	P@10	P@20	SSIM
8	0.40	0.80	0.60	0.20	0.863 7
16	0.63	1.00	0.70	0.50	0.891 4
32	0.68	1.00	0.80	0.55	0.893 9
64	0.71	1.00	0.90	0.55	0.904 3
128	0.00	0.00	0.00	0.00	0.414 5
256	0.00	0.00	0.00	0.00	0.365 8

主干层	AP	P@5	P@10	P@20	SSIM
深度卷积网络	0.46	0.6	0.5	0.4	0.893 5
残差网络	0.71	1.00	0.9	0.55	0.904 3

指标	算法
指标	传统TSA	多假设TSA	multi-frame S-D TSA	本文方法
AP/%	42.8	94.0	97.5	100.0
t/s	0.35	0.38	1.25	0.12

噪声等级	AP	P@5	P@10	P@20	SSIM
2	0.71	1.00	0.9	0.55	0.904 3
4	0.69	1.00	0.8	0.45	0.873 6
6	0.58	1.00	0.7	0.4	0.834 2

[1]	韩啸, 陈世文, 陈蒙, 杨锦程. 基于互易点学习的LPI信号开集识别[J]. 系统工程与电子技术, 2022, 44(9): 2752-2759.
[2]	刘戎翔, 吴琳, 谢智歌, 刘虹麟. 基于生成对抗网络的防空体系态势辅助分析[J]. 系统工程与电子技术, 2022, 44(8): 2522-2529.
[3]	曹鹏宇, 杨承志, 石礼盟, 吴宏超. 基于PSO-DBSCAN和SCGAN的未知雷达信号处理方法[J]. 系统工程与电子技术, 2022, 44(4): 1158-1165.
[4]	邵凯, 朱苗苗, 王光宇. 基于生成对抗与卷积神经网络的调制识别方法[J]. 系统工程与电子技术, 2022, 44(3): 1036-1043.
[5]	吴涛, 王伦文, 朱敬成. 基于迁移学习和注意力机制的伪装图像分割[J]. 系统工程与电子技术, 2022, 44(2): 376-384.
[6]	金涛, 王晓峰, 田润澜, 张歆东. 基于改进1DCNN+TCN的雷达辐射源快速识别方法[J]. 系统工程与电子技术, 2022, 44(2): 463-469.
[7]	马钰棠, 孙鹏, 张杰勇, 王鹏, 闫云飞, 赵亮. 样本不平衡下的空中群组意图识别方法[J]. 系统工程与电子技术, 2022, 44(12): 3747-3755.
[8]	唐一强, 杨霄鹏, 朱圣铭. 基于注意力机制的混合CNN-BiLSTM低轨卫星信道预测算法[J]. 系统工程与电子技术, 2022, 44(12): 3863-3870.
[9]	曲凌志, 杨俊安, 刘辉, 黄科举. 嵌入注意力机制的通信辐射源个体识别方法[J]. 系统工程与电子技术, 2022, 44(1): 20-27.
[10]	刘紫燕, 马珊珊, 梁静, 朱明成, 袁磊. 注意力机制CNN的毫米波大规模MIMO系统信道估计算法[J]. 系统工程与电子技术, 2022, 44(1): 307-312.
[11]	曹鹏宇, 杨承志, 石礼盟, 吴宏超. 基于DAE-GAN网络的LPI雷达信号增强[J]. 系统工程与电子技术, 2021, 43(9): 2493-2500.
[12]	崔邦彦, 田润澜, 王东风, 崔钢, 石静苑. 基于注意力机制和改进CLDNN的雷达辐射源识别[J]. 系统工程与电子技术, 2021, 43(5): 1224-1231.
[13]	赵凡, 金虎. 基于GAN的通信干扰波形生成技术[J]. 系统工程与电子技术, 2021, 43(4): 1080-1088.
[14]	高诗飏, 董会旭, 田润澜, 张歆东. 基于SRNN+Attention+CNN的雷达辐射源信号识别方法[J]. 系统工程与电子技术, 2021, 43(12): 3502-3509.
[15]	赵若辰, 王敬东, 林思玉, 顾东泽. 基于卷积神经网络的小型建筑物检测算法[J]. 系统工程与电子技术, 2021, 43(11): 3098-3106.