基于高分辨率孪生网络的无人机目标跟踪算法

doi:10.13229/j.cnki.jdxbgxb.20220790

摘要/Abstract

摘要：

针对无人机（UAV）目标跟踪任务中目标尺寸小、尺度变化明显和视点改变频繁等问题，本文提出一种基于高分辨率孪生网络的无人机目标跟踪算法。首先，利用改进高分辨率网络作为特征提取主干网络，并且采用动态多模板策略挖掘视频的帧间信息；然后，构建多帧特征融合模块，得到利于目标定位的融合特征；最后，选取无锚框策略定位目标位置，得到精确的跟踪结果。实验结果表明：本文算法在DTB70数据集测试的成功率和准确率分别为66.0%和84.7%，在UAV123数据集测试的成功率和准确率分别为65.7%和84.3%，有效地提升了目标跟踪性能。

关键词: 计算机视觉, 目标跟踪, 无人机, 高分辨率孪生网络, 多帧特征

Abstract:

Unmanned Aerial Vehicle（UAV） target tracking tasks are often suffer from small target size， large scale variance and frequent viewpoint change. To address these issues， in this paper， an UAV target tracking algorithm based on high-resolution siamese network is proposed. Firstly， a high-resolution network is improved as the feature extraction backbone network （Lite-high resolution network， L-HRNet）， and a dynamic multi-template strategy is used to mine the inter-frame information of the video. Secondly， a multi-frame feature fusion module is constructed to obtain fusion features that are beneficial to target localization. Finally， an anchor-free strategy is selected to locate the target position and obtain accurate tracking results. The experimental results show that the success rate and accuracy of the proposed algorithm are 66.0% and 84.7% on DTB70 dataset， 65.7% and 84.3% on UAV123 dataset respectively， which improves the target tracking performance effectively.

Key words: computer vision, target tracking, unmanned aerial vehicle, high-resolution siamese network, multi-frame feature

中图分类号:

TP391.4

王殿伟,张池,房杰,许志杰. 基于高分辨率孪生网络的无人机目标跟踪算法[J]. 吉林大学学报(工学版), 2024, 54(5): 1426-1434.

Dian-wei WANG,Chi ZHANG,Jie FANG,Zhi-jie XU. UAV target tracking algorithm based on high resolution siamese network[J]. Journal of Jilin University(Engineering and Technology Edition), 2024, 54(5): 1426-1434.

图/表 10

图1

图2

图3

表1

图4

图5

表2

图6

表3

表4

参考文献 29

1	谢才华, 王学伟, 付强, 等. 基于动态模板更新的孪生网络目标跟踪算法[J]. 吉林大学学报: 工学版, 2022, 52(5): 1106-1116.
	Xie Cai-hua, Wang Xue-wei, Fu Qiang, et al. Siamese network target tracking algorithm based on dynamic template update[J]. Journal of Jilin University (Engineering Edition), 2022, 52(5): 1106-1116.
2	刘晶红, 邓安平, 陈琪琪, 等. 基于多重注意力机制的无锚框目标跟踪算法[J]. 吉林大学学报:工学版, 2023, 53(12): 3518-3528.
	Liu Jing-hong, Deng An-ping, Chen Qi-qi, et al. Anchor-free target tracking algorithm based on multiple attention mechanism[J]. Journal of Jilin University (Engineering and Technology Edition), 2023, 53(12): 3518-3528.
3	Bertinetto L, Valmadre J, Henriques J F, et al. Fully convolutional Siamese networks for object tracking[C]∥Europe an Conference on Computer Vision. Berlin: Springer, 2016: 850-865.
4	Li B, Yan J, Wu W, et al. High performance visual tracking with Siamese region proposal network[C]∥Proceedings. of the IEEE Conference on Computer Vision and Pattern Recognition, Utah, USA, 2018: 8971-8980.
5	Li B, Wu W, Wang Q, et al. SiamRPN++: evolution of Siamese visual tracking with very deep networks[C]∥Proceedings. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, California, USA, 2019: 4282-4291.
6	He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]∥Proceedings. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Nevada,USA,2016: 770 -778.
7	Guo D, Wang J, Cui Y, et al. SiamCAR: Siamese fully convolutional classification and regression for visual tracking[C]∥Proceedings. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 6269-6277.
8	Chen Z, Zhong B, Li G, et al. Siamese box adaptive network for visual tracking [C]∥Proceedings. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 6668-6677.
9	Xu Y, Wang Z, Li Z, et al. SiamFC++: towards robust and accurate visual tracking with target estimation guidelines[C]∥Proceedings. of the AAAI Conference on Artificial Intelligence, 2020, 34(7): 12549-12556.
10	Han K, Wang Y, Chen H, et al. A survey on visual transformer[J]. arXiv preprint arXiv: 2012.12556, 2020, 2(4).
11	Khan S, Naseer M, Hayat M, et al. Transformers in vision: a survey[J]. ACM Computing Surveys (CSUR), 2022, 54(10s): 1-41.
12	Chen X, Yan B, Zhu J, et al. Transformer tracking[C]∥Proceedings. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021: 8126-8135.
13	Wang N, Zhou W, Wang J, et al. Transformer meets tracker: exploiting temporal context for robust visual tracking [C]∥Proceedings. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021:1571-1580.
14	Fu Z, Liu Q, Fu Z, et al. Stmtrack: template-free visual tracking with space time memory networks[C]∥Proceedings. of the IEEE /CVF Conference on Computer Vision and Pattern Recognition, 2021: 13774-13783.
15	Zhao M, Okada K, Inaba M. TrTr: visual tracking with transformer[J]. arXiv preprint arXiv: 2105. 03817, 2021.
16	Cui Y, Jiang C, Wang L, et al. MixFormer: end-to-end tracking with iterative mixed attention[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022: 13608-13618.
17	Cao Z, Huang Z, Pan L, et al. TCTrack: temporal contexts for aerial tracking[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition,2022: 14798-14808.
18	Song Z, Yu J, Chen Y P, et al. Transformer tracking with cyclic shifting window attention[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022: 8791-8800.
19	Cao Z, Fu C, Ye J, et al. HiFT: hierarchical feature transformer for aerial tracking[C]∥Proceedings of the IEEE/CVF International Conference on Computer Vision, 2021: 15457-15466.
20	Fu C, Li S, Yuan X, et al. Ad2Attack: adaptive adversarial attack on real-time UAV tracking [J]. arXiv preprint arXiv:, 2022.
21	Xing D, Evangeliou N, Tsoukalas A, et al. Siamese transformer pyramid networks for real-time UAV tracking[C]∥Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2022: 2139-2148.
22	Simonyan K, Zisserman A. Very deep convolutional networks for large scale image recognition[J]. arXiv: , 2014.
23	Krizhevsky A, Sutskever I, Hinton G E. Imagenet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.
24	Sun K, Xiao B, Liu D, et al. Deep high-resolution representation learning for human pose estimation[C]∥Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, California, USA, 2019: 5693-5703.
25	Zhang Y, Wang C, Wang X, et al. Fairmot: on the fairness of detection and reidentification in multiple object tracking[J]. International Journal of Computer Vision, 2021, 129(11): 3069-3087.
26	Russakovsky O, Deng J, Su H, et al. Imagenet large scale visual recognition challenge[J]. International Journal of Computer Vision, 2015, 115(3): 211-252.
27	Huang L, Zhao X, Huang K. Got-10k: a large high-diversity benchmark for generic object tracking in the wild[J]. IEEE Trans. on Pattern Analysis and Machine Intelligence, 2019, 43(5): 1562-1577.
28	Li S, Yeung D Y. Visual object tracking for unmanned aerial vehicles: a benchmark and new motion models[C]∥Thirty-first AAAI Conference on Artificial Intelligence, San Francisco, USA, 2017.
29	Mueller M, Smith N, Ghanem B. A benchmark and simulator for UAV tracking [C]∥European Conference on Computer Vision. Belin: Springer, 2016: 445-446.

相关文章 15

[1]	王宇,赵凯. 基于亚像素定位的人体姿态热图后处理[J]. 吉林大学学报(工学版), 2024, 54(5): 1385-1392.
[2]	高云龙,任明,吴川,高文. 基于注意力机制改进的无锚框舰船检测模型[J]. 吉林大学学报(工学版), 2024, 54(5): 1407-1416.
[3]	陈小波,陈玲. 定位噪声统计特性未知的变分贝叶斯协同目标跟踪[J]. 吉林大学学报(工学版), 2023, 53(4): 1030-1039.
[4]	胡江平,周子粲,陈波. 基于方位测量的固定时间多目标定位和环航控制[J]. 吉林大学学报(工学版), 2023, 53(3): 923-932.
[5]	徐卓君,王耀祥,黄兴,彭程. 多无人机地面移动目标搜寻和定位[J]. 吉林大学学报(工学版), 2023, 53(3): 832-840.
[6]	申富媛,李炜,蒋栋年. 四旋翼无人机寿命预测和自主维护方法[J]. 吉林大学学报(工学版), 2023, 53(3): 841-852.
[7]	于雅静,郭健,王荣浩,秦伟,宋明武,向峥嵘. 基于状态观测器的多四旋翼无人机时变编队控制[J]. 吉林大学学报(工学版), 2023, 53(3): 871-882.
[8]	薛珊,张亚亮,吕琼莹,曹国华. 复杂背景下的反无人机系统目标检测算法[J]. 吉林大学学报(工学版), 2023, 53(3): 891-901.
[9]	陈珑茏,冯天宇,吕宗阳,吴玉虎. 共轴倾转旋翼无人机有限时间滑模姿态控制[J]. 吉林大学学报(工学版), 2023, 53(3): 883-890.
[10]	何颖,樊俊松,王巍,孙庚,刘衍珩. 无人机空地安全通信与航迹规划的多目标联合优化方法[J]. 吉林大学学报(工学版), 2023, 53(3): 913-922.
[11]	王小艺,刘迪一,于家斌,何卓昀,赵峙尧. 复杂风场环境下的多旋翼无人机编队故障检测方法[J]. 吉林大学学报(工学版), 2023, 53(3): 823-831.
[12]	潘弘洋,刘昭,杨波,孙庚,刘衍珩. 基于新一代通信技术的无人机系统群体智能方法综述[J]. 吉林大学学报(工学版), 2023, 53(3): 629-642.
[13]	刘晶红,邓安平,陈琪琪,彭佳琦,左羽佳. 基于多重注意力机制的无锚框目标跟踪算法[J]. 吉林大学学报(工学版), 2023, 53(12): 3518-3528.
[14]	曲优,李文辉. 基于多任务联合学习的多目标跟踪方法[J]. 吉林大学学报(工学版), 2023, 53(10): 2932-2941.
[15]	鲜斌,李杰奇,古训. 基于非线性扰动观测器的无人机地面效应补偿[J]. 吉林大学学报(工学版), 2022, 52(8): 1926-1933.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

算法	成功率	精确率
SiamFC	0.487	0.738
SiamRPN	0.588	0.790
DaSiamRPN	0.610	0.796
SiamRPN++	0.614	0.800
SiamCAR	0.596	0.802
SiamBAN	0.643	0.834
本文	0.660	0.847

算法	低分辨率		尺度变化		视点改变		纵横比改变		相似物体		全遮挡
算法	S	P	S	P	S	P	S	P	S	P	S	P
SiamFC	0.346	0.719	0.476	0.697	0.479	0.686	0.436	0.661	0.466	0.706	0.306	0.571
SiamRPN	0.400	0.616	0.544	0.752	0.608	0.779	0.536	0.724	0.487	0.682	0.349	0.565
DaSiamRPN	0.411	0.663	0.560	0.754	0.563	0.753	0.537	0.739	0.517	0.747	0.379	0.633
SiamRPN++	0.433	0.666	0.578	0.775	0.637	0.833	0.544	0.756	0.541	0.740	0.354	0.582
SiamCAR	0.465	0.693	0.605	0.791	0.646	0.807	0.580	0.759	0.563	0.754	0.419	0.660
SiamBAN	0.472	0.719	0.613	0.813	0.639	0.824	0.591	0.796	0.566	0.777	0.418	0.671
本文	0.494	0.734	0.641	0.824	0.690	0.864	0.634	0.784	0.571	0.762	0.415	0.658

骨干网络	成功率/%	精确率/%
AlexNet	61.3	78.5
HRNet	63.0	81.7
L-HRNet	65.7	84.3

模板数量	成功率/%	精确率/%	帧率/（帧·s^-1）
1	61.5	80.6	46
2	64.8	83.0	34
3	65.7	84.3	29