基于滤波器自适应更新的机场目标跟踪算法

杨临风; 牟睿; 黎新; 李炜

doi:10.3963/j.jssn.1674-4861.2022.01.009

基于滤波器自适应更新的机场目标跟踪算法

doi: 10.3963/j.jssn.1674-4861.2022.01.009

杨临风^{1, 2, 3,},
牟睿^4, ,,
黎新⁵,
李炜⁶

1.
中国民用航空飞行学院民机火灾科学与安全工程四川省重点实验室四川广汉 618307
2.
四川川大智胜软件股份有限公司成都 610045
3.
中国民用航空飞行学院空中交通管理学院四川广汉 618307
4.
中国民用航空飞行学院广汉分院四川广汉 618307
5.
中国民用航空飞行学院民航安全工程学院四川广汉 618307
6.
四川大学空天科学与工程学院成都 610065

基金项目:

国家重点研发计划项目 2018YFC0809503

四川省科技计划重点研发项目 2020YFG0134

详细信息

作者简介:
杨临风（1993—），硕士. 研究方向：空中交通管理. E-mail: 1002120943@qq.com

通讯作者:
牟睿（1993—），本科，三级飞行员. 研究方向：飞行技术及智慧机场.E-mail: 285921595@qq.com

中图分类号: TP391.41
计量
- 文章访问数: 1222
- HTML全文浏览量: 552
- PDF下载量: 29
- 被引次数: 0
出版历程
- 收稿日期: 2021-08-11
- 网络出版日期: 2022-03-31

Development of an Object Tracking Algorithm for Airports Using Adaptive Filter Update Technique

YANG Linfeng^{1, 2, 3
,},
MOU Rui^{4
, ,},
LI Xin⁵,
LI Wei⁶

1.
Civil Aircraft Fire Science and Safety Engineering Key Laboratory of Sichuan Province, Civil Aviation Flight University of China, Guanghan 618307, Sichuan, China
2.
WisesoftCo., Ltd., Chengdu 610045, China
3.
College of Air Traffic Management, Civil Aviation Flight University of China, Guanghan 618307, Sichuan, China
4.
Civil Aviation Flight University of China Guang Han College, Guanghan 618307, Sichuan, China
5.
College of Civil Aviation Safety Engineering, Civil Aviation Flight University of China, Guanghan 618307, Sichuan, China
6.
School of Aeronautics and Astronautics, Sichuan University, Chengdu 610065, China

摘要

摘要: 机场场面目标跟踪常面临目标遮挡、背景干扰、低分辨率等因素的影响，导致跟踪准确性降低甚至丢失跟踪目标。针对以上问题，研究了基于滤波器自适应更新的机场目标跟踪算法。选取跟踪目标的颜色特征和深度特征，通过插值算子进行多特征融合，再将融合特征与之对应的滤波器进行卷积求和计算各区域置信度，置信度高的区域即为跟踪目标位置。为提高跟踪准确性，利用峰值旁瓣比与平均响应峰值能量建立了跟踪结果校验机制，并设计了1种滤波器自适应更新策略，使滤波器能够自适应调整学习速率，仅在结果可靠时更新。在西南某机场采集的视频数据集上进行测试，结果表明：算法在目标特征不明显或发生变化时具有更好的性能，在目标遮挡和背景干扰等9种因素下的跟踪性能有较大提升，整体精确度和成功率分别达到0.834和0.828，较原ECO算法分别提升了11.35%和11.29%，且均优于文中提到的其他5种经典算法。
- 智能交通 /
- 场面监视 /
- 自适应更新 /
- 相关滤波 /
- 卷积神经网络 /
- 多特征融合
Abstract: Tracking objects over airport surface is often hindered by the factors such as occlusion, background clutter and low resolution, which often result in reduced tracking accuracy or even loss of tracked objects. In order to mitigate the above problems, an object tracking algorithm for airports based on adaptive filter update is developed. First, the color and convolutional neural network feature of the tracking object are selected. Based on these features, multi-feature fusion is performed through an interpolation operator. Then, the fusion feature and its corresponding filter are convolved and summedin order to calculate the confidence level of each region.Theregion with a high confidence level is then identified as thelocation of the tracked object. By using the peak to side-lobe ratio and the average peak-to-correlation energy, a verification method is developed to improve the tracking accuracy. Furthermore, a self-adaptive updating algorithm is designed to adjust the learning rate of the filter and updated only when the results are reliable. According to the results obtained using a video dataset collected at an airport in Southwest China, the proposed algorithm has a better tracking performance when the object features change or are unclear, and the results also indicate the tracking performance is significantly improved under 9 different factors, such as occlusions and background clutter. The overall accuracy and success rate are 0.834 and 0.828 respectively, which are higher than that of the original ECO algorithm by 11.35% and 11.29%, and are superior to the other five classical algorithms.
- intelligent transportation systems /
- surface surveillance /
- adaptive update /
- correlation filter /
- convolutional neural network /
- multi-feature fusion

HTML全文

图 1 ECO算法流程图

Figure 1. ECO algorithm flow chart

下载: 全尺寸图片幻灯片

图 2 可视化特征图

Figure 2. Visualized feature map

下载: 全尺寸图片幻灯片

图 3 本文算法流程图

Figure 3. Proposed algorithm flow chart

下载: 全尺寸图片幻灯片

图 4 遮挡因素下的响应图变化

Figure 4. Response map variation in occlusion factor

下载: 全尺寸图片幻灯片

图 5 2种校验机制的测试结果

Figure 5. Test results of the two verification mechanisms

下载: 全尺寸图片幻灯片

图 6 本文算法与其他算法的测试结果

Figure 6. Test results of the proposed and other algorithms

下载: 全尺寸图片幻灯片

图 7 本文和其他算法的跟踪效果

Figure 7. Tracking effect of the proposed and other algorithms

下载: 全尺寸图片幻灯片

表 1 跟踪因素表

Table 1. Tracking factors table

序号	跟踪因素
1	光照变化(illumination variation, IV)
2	尺度变化(scale variation, SV)
3	平面外旋转(out-of-plane rotation, OPR)
4	遮挡（occlusion, OCC)
5	变形(deformation, DEF)
6	运动模糊（motionblur, MB)
7	快速运动（fast motion, FM)
8	平面内旋转(in-plane rotation, IPR)
9	出视野(out-of-view, OV)
10	背景干扰(backgroundclutters, BC)
11	低分辨率(lowresolution，LR)

下载: 导出CSV

表 2 不同跟踪因素下的精确度对比表

Table 2. Precision comparison table under different tracking factors

跟踪因素	ECO	本文算法
IV	0.626	0.705
SV	0.733	0.785
OPR	0.782	0.884
OCC	0.779	0.833
DEF	0.703	0.862
MB	0.620	0.698
FM	0.694	0.667
IPR	0.689	0.833
OV	0.833	0.765
BC	0.619	0.765
LR	0.611	0.701
Overall	0.749	0.834

下载: 导出CSV

表 3 不同跟踪因素下的成功率对比表

Table 3. Success rates comparison table under different tracking factors

跟踪因素	ECO	本文算法
IV	0.602	0.695
SV	0.757	0.798
OPR	0.716	0.847
OCC	0.769	0.823
DEF	0.724	0.872
MB	0.643	0.709
FM	0.673	0.662
IPR	0.681	0.789
OV	0.830	0.777
BC	0.624	0.762
LR	0.610	0.686
Overall	0.744	0.828

下载: 导出CSV

参考文献(17)

[1]	高策, 褚端峰, 何书贤, 等. 基于卡尔曼-高斯联合滤波的车辆位置跟踪[J]. 交通信息与安全, 2020, 38(1): 76-83. doi: 10.3963/j.jssn.1674-4861.2020.01.010 GAO C, CHU D F, HE S X, et al. Vehicle position tracking based on joint kalman-gaussian filter[J]. Journal of Transport Information and Safety, 2020, 38(1): 76-83. (in Chinese) doi: 10.3963/j.jssn.1674-4861.2020.01.010
[2]	梁海军, 张雪华, 夏正洪. 基于概率修正的机场场面运动航空器跟踪算法[J]. 科学技术与工程, 2017, 17(19): 84-91. doi: 10.3969/j.issn.1671-1815.2017.19.014 LIANG H J, ZHANG X H, XIA Z H. Tracking algorithm of moving aircraft in airport based on probability correction[J]. Science Technology and Engineering, 2017, 17(19): 84-91. (in Chinese) doi: 10.3969/j.issn.1671-1815.2017.19.014
[3]	潘卫军, 罗杰, 王少杰, 等. 基于跑滑系统约束的航空器滑行跟踪算法[J]. 四川大学学报(自然科学版), 2019, 56(5): 843-850. doi: 10.3969/j.issn.0490-6756.2019.05.009 PAN W J, LUO J, WANG S J, et al. Taxiing tracking algorithm of aircraft on the ground based on the runway-taxiway system constraints[J]. Journal of Sichuan University(Natural Science Edition), 2019, 56(5): 843-850. (in Chinese) doi: 10.3969/j.issn.0490-6756.2019.05.009
[4]	穆柯楠, 王会峰, 杨澜, 等. 基于多尺度边缘融合及SURF特征匹配的车辆检测及跟踪方法[J]. 交通信息与安全, 2018, 36(6): 65-73. doi: 10.3963/j.issn.1674-4861.2018.06.009 MU K N, WANG H F, YANG L, et al. A method of detection and tracking vehicles based on multi-scale edge fusion and SURF feature matching[J]. Journal of Transport Information and Safety, 2018, 36(6): 65-73. (in Chinese) doi: 10.3963/j.issn.1674-4861.2018.06.009
[5]	孙博, 王阿川. 融合深度特征和FHOG特征的尺度自适应相关滤波跟踪算法[J]. 河北科技大学学报, 2021, 42(6): 591-600. https://www.cnki.com.cn/Article/CJFDTOTAL-JGDJ202004036.htm SUN B, WANG A C. Scale-adaptive correlation filter tracking algorithm fusing depth features and FHOG features[J]. Journal of Hebei University of Science and Technology, 2021, 42 (06): 591-600. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JGDJ202004036.htm
[6]	孟晓燕, 段建民. 基于相关滤波的目标跟踪算法研究综述[J]. 北京工业大学学报, 2020, 46(12): 1393-1416. doi: 10.11936/bjutxb2019030011 MENG X Y, DUAN J M. Advances in correlation filter-based object tracking algorithms: a review[J]. Journal of Beijing University of Technology, 2020, 46(12): 1393-1416. (in Chinese) doi: 10.11936/bjutxb2019030011
[7]	徐宁, 王娟娟, 郭晓雨, 等. 判别式相关滤波器的目标跟踪综述[J]. 小型微型计算机系统, 2020, 41(12): 2484-2493. doi: 10.3969/j.issn.1000-1220.2020.12.006 XU N, WANG J J, GUO X Y, et al. Survey of visual tracking based on discriminative correlation filters[J]. Journal of Chinese Computer Systems, 2020, 41(12): 2484-2493. (in Chinese) doi: 10.3969/j.issn.1000-1220.2020.12.006
[8]	WANG M M, LIU Y, HUANG Z Y. Large margin object tracking with circulant feature maps[C]. IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA: IEEE, 2017.
[9]	谢维信, 赵田. 多特征自适应融合的相关滤波目标跟踪算法[J]. 信号处理, 2021, 37(4): 603-615. https://www.cnki.com.cn/Article/CJFDTOTAL-XXCN202104015.htm XIE W X, ZHAO T. Multi-feature adaptive fusion based target tracking algorithm[J]. Journal of Signal Processing, 2021, 37(4): 603-615. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-XXCN202104015.htm
[10]	李健宁, 曹文君, 刘晓利. 结合多尺度改进颜色特征应对遮挡的跟踪算法[J]. 兵器装备工程学报, 2019, 40(6): 134-138+197. doi: 10.11809/bqzbgcxb2019.06.028 LI J N, CAO W J, LIU X L. Tracking algorithm for target occlusion by improving color features combined with scale pyramid[J]. Journal of Ordnance Equipment Engineering, 2019, 40(6): 134-138+197. (in Chinese) doi: 10.11809/bqzbgcxb2019.06.028
[11]	王科平, 朱朋飞, 杨艺, 等. 多时空感知相关滤波融合的目标跟踪算法[J]. 计算机辅助设计与图形学学报, 2020, 32 (11): 1840-1852. https://www.cnki.com.cn/Article/CJFDTOTAL-JSJF202011014.htm WANG K P, ZHU P F, YANG Y, et al. Target tracking algorithm based on multi-time-space perception correlation filters fusion[J]. Journal of Computer-Aided Design and Computer Graphics, 2020, 32(11): 1840-1852. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JSJF202011014.htm
[12]	毛宁, 杨德东, 杨福才, 等. 基于分层卷积特征的自适应目标跟踪[J]. 激光与光电子学进展, 2016, 53(12): 195-207. https://www.cnki.com.cn/Article/CJFDTOTAL-JGDJ201612027.htm MAO N, YANG D D, YANG F C, et al. Adaptive object tracking based on hierarchical convolution features[J]. Laser and Optoelectronics Progress, 2016, 53(12): 195-207. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-JGDJ201612027.htm
[13]	MA C, HUANG J B, YANG X K, et al. Hierarchical convolutional features for visual tracking[C]. IEEE International Conference on Computer Vision, Santiago, Chile: IEEE, 2015.
[14]	DANELLJAN M, BHAT G, KHAN F S, et al. ECO: Efficient convolution operators for tracking[C]. IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA: IEEE, 2017.
[15]	BHAT G, JOHNANDER J, DANELLJAN M. Unveiling the power of deep tracking[C]. European Conference on Computer Vision, Munich, Germany: Springer, 2018.
[16]	李欣, 周婧琳, 厚佳琪, 等. 基于ECO-HC改进的运动目标跟踪方法研究[J]. 南京大学学报(自然科学), 2020, 56(2): 216-226. https://www.cnki.com.cn/Article/CJFDTOTAL-NJDZ202002007.htm LI X, ZHOU J L, HOU J Q, et al. Research on improved moving object tracking method based on ECO-HC[J]. Journal of Nanjing University(Natural Science), 2020, 56(2): 216-226. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-NJDZ202002007.htm
[17]	李国友, 张凤熙, 纪执安. 自适应多滤波器的高效卷积算子目标跟踪算法[J]. 光电工程, 2020, 47(7): 50-62. https://www.cnki.com.cn/Article/CJFDTOTAL-GDGC202007006.htm LI G Y, ZHANG F X, JI Z A. Adaptive multi-filter tracker based on efficient convolution operator[J]. Opto-Electronic Engineering, 2020, 47(7): 50-62. (in Chinese) https://www.cnki.com.cn/Article/CJFDTOTAL-GDGC202007006.htm