A Combined Feature-based Algorithm of Target Detection and Tracking Used

doi:10.12382/bgxb.2022.0017

Abstract

Abstract:

To deal with the difficulty of underwater artificial target detection and tracking via three-dimensional (3D) imaging sonar, a combined geometric feature-based underwater 3D target detection and tracking method is proposed. The algorithm uses distribution information such as the intensity and depth of the sonar image, and the background difference method to extract the kernel function according to the changes in the energy of the sonar image before and after the target appears. Besides, the detection and tracking of the artificial target using three-dimensional mapping instead of two-dimensional mapping is realized by correlating combined geometric features including the Hu invariant moment, target centroid and target distance of the previous and subsequent frame images on the basis of the correlation matching of the two-dimensional sonar image. The data processing results of the lake test show that the algorithm improves tracking performance by about 7.15%, and consumes about 60% less time than the traditional method, which verifies the effectiveness of the algorithm.

Key words: 3D imaging sonar, detection and tracking, combined geometric feature, template matching

ZENG Teng, REN Lulu, WANG Yujie, WANG Peng, HUANG Haining. A Combined Feature-based Algorithm of Target Detection and Tracking Used[J]. Acta Armamentarii, 2023, 44(5): 1384-1393.

Figures/Tables 18

Fig.1 Working diagram of 3D imaging sonar work

Fig.2 Diagram of 3D imaging sonar beam energy per frame

Fig.3 Diagram of 3D sonar beam direction

Fig.4 Diagram of beam envelope bb(t, $\hat{u}$)

Fig.5 Diagram of intensity and distance

Fig.6 Diagram of target detection based on background elimination method

Fig.7 Diagram of template matching principle

Fig.8 Slice energy distribution at the maximum filtering distance

Fig.9 Diagram of first frame initialization

Fig.10 Diagram of intensity image tracking based on combined features

Fig.11 Diagram of 3D detection box mapping

Fig.12 Test target and site

Fig.13 Diagram of first frame sonar intensity of steel-frame target

Fig.14 Second frame detection and tracking results

Fig.15 Comparison of overlap rate of different tracking methods

Fig.16 Comparison of different μ vaules

Table 1 Factors in the table

μ	3	5	8	10	15	20
t/s	102.82	77.53	58.83	58.94	36.08	31.05

Fig.17 Comparison of different F values

References 20

[1]	刘立昕. 成像声呐目标检测与跟踪技术研究[D]. 哈尔滨: 哈尔滨工程大学, 2015.
	LIU L X. Research on target detection and tracking technology using imaging sonar[D]. Harbin:Harbin Engineering University, 2015. (in Chinese)
[2]	卞红雨, 陈奕名, 张志刚, 等. 像素重要性测量特征下的侧扫声呐目标检测[J]. 声学学报, 2019, 44(3):353-359.
	BIAN H Y, CHEN Y M, ZHANG Z G, et al. Target detection algorithm in side-scan sonar image based on pixel importance measurement[J]. Chinese Journal of Acoustics, 2019, 44(3):353-359. (in Chinese)
[3]	LIPTON A J, FUJIYOSHI H, PATIL R S. Moving target classification and tracking from real-time video[C]∥Proceedings of the 4th IEEE Workshop on Applications of Computer.Princeton,PA, US:IEEE, 1998:8-14.
[4]	VALERA M, VELASTIN S A. Intelligent distributed surveillance systems: a review[J]. IEEE Proceeding of Vision, Image and Signal Processing, 2005, 152(2):192-204.
[5]	甘明刚, 陈杰, 刘劲, 等. 一种基于三帧差分和边缘信息的运动目标检测方法[J]. 电子与信息学报, 2010, 32(4):894-897.
	GAN M G, CHEN J, LIU J, et al. Moving object detection algorithm based on three-frame-differencing and edge information[J]. Journal of Electronics & Information Technology, 2010, 32(4): 894-897. (in Chinese)
[6]	李毅, 孙正兴, 远博. 一种改进的帧差和背景减相结合的运动检测方法[J]. 中国图象图形学报, 2009, 14(6): 1162-1168.
	LI Y, SUN Z X, YUAN B. An improved method for motion detection by frame difference and background subtraction[J]. Journal of Image and Graphics, 2009, 14(6):1162-1168. (in Chinese)
[7]	李玉钰, 颜景龙. 多特征联合匹配的目标图像稳定跟踪算法[J]. 兵工学报, 2011, 32(5):574-579.
	LI Y Y, YAN J L. Stabilized target tracking algorithm using multi-feature joint matching[J]. Acta Armamentarii, 2011, 32(5):574-579. (in Chinese)
[8]	孟琭, 杨旭. 目标跟踪算法综述[J]. 自动化学报, 2019, 45(7):1244-1260.
	MENG L, YANG X. A survey of object tracking algorithms[J]. IEEE/CAA Journal of Automatica Sinica, 2019, 45(7):1244-1260. (in Chinese)
[9]	LI C M, XU C Y, GUI C F, et al. Distance regularized level set evolution and its application to image segmentation[J]. IEEE Transactions on Image Processing, 2010, 19(12):3243-3253. doi: 10.1109/TIP.2010.2069690 pmid: 20801742
[10]	ARNOLD E, AL-JARRAH O Y, DIANATI M, et al. A survey on 3D object detection methods for autonomous driving applications[J]. IEEE Transactions on Intelligent Transportation Systems, 2019, 20(10):3782-3795. doi: 10.1109/TITS.6979 URL
[11]	王朋. 基于稀疏布阵的三维成像声呐信号处理算法研究[D]. 北京: 中国科学院大学, 2015.
	WANG P. Research on signal processing algorithm of three-dimensional acoustical imaging sonar based on sparse planar array[D]. Beijing: University of Chinese Academy of Sciences, 2015. (in Chinese)
[12]	曾腾. 三维成像声呐图像重建与显示技术研究[D]. 北京: 中国科学院大学, 2020.
	ZENG T. Research on image reconstruction and display technology of 3-D imaging sonar system[D]. Beijing: University of Chinese Academy of Sciences, 2020. (in Chinese)
[13]	MURINO V. Three-dimensional image generation and processing in underwater acoustic vision[J]. Proceeding of the IEEE, 2000, 88(12):1903-1948. doi: 10.1109/5.899059 URL
[14]	HOUSTON K M. Three-dimensional acoustical imaging using micromechanical hydrophones[C]∥Proceedings of Oceans MTS/IEEE.San Diego, CA, US:IEEE, 1995:1174-1182.
[15]	董静. 基于图像声呐的动目标检测技术研究[D]. 哈尔滨: 哈尔滨工程大学, 2014.
	DONG J. Research on moving targets detection based on image sonar[D]. Harbin:Harbin Engineering University, 2014. (in Chinese)
[16]	阮秋琦. 数字图像处理[M]. 第3版. 北京: 电子工业出版社, 2011:560-564.
	RUAN Q Q. Digital image processing, third edition[M]. 3rd Edition. Beijing: Publishing House of Electronics Industry, 2011:560-564. (in Chinese)
[17]	付萍, 薛定宇, 林明秀, 等. 基于单目视觉的多车辆跟踪方法研究[J]. 兵工学报, 2007, 28(8):988-992.
	FU P, XUE D Y, LIN M X, et al. Research on tracking method of multiple vehicles based on monocular camera[J]. Acta Armamentarii, 2007, 28(8):988-992. (in Chinese)
[18]	曾腾, 张春华, 王朋. 基于局部异常因子算法的三维声呐单帧重建研究[J]. 兵工学报, 2020, 41(3):552-558. doi: 10.3969/j.issn.1000-1093.2020.03.016
	ZENG T, ZHANG C H, WANG P. Research on the 3D sonar single-frame reconstruction based on local outlier factor algorithm[J]. Acta Armamentarii, 2020, 41(3):552-558. (in Chinese) doi: 10.3969/j.issn.1000-1093.2020.03.016
[19]	姚玉荣, 章毓晋. 利用小波和矩进行基于形状的图象检索[J]. 中国图象图形学报, 2000, 5(3): 206-210.
	YAO Y R, ZHANG Y J. Shape-based image retrieval using wavelet and moment[J]. Journal of Image and Graphics, 2000, 5(3): 206-210. (in Chinese)
[20]	孙彦景, 王赛楠, 石韫开, 等. 基于全局背景与特征降维的视觉跟踪算法[J]. 电子与信息学报, 2018, 40(9):2135-2142.
	SUN Y J, WANG S N, SHI Y K, et al. Visual tracking algorithm based on global context and feature dimensionality reduction[J]. Journal of Electronics & Information Technology, 2018, 40(9):2135-2142. (in Chinese)