复杂动背景下的“低小慢”目标检测技术

吴言枫; 王延杰; 孙海江; 刘培勋

doi:10.3788/CO.20191204.0853

复杂动背景下的“低小慢”目标检测技术

doi: 10.3788/CO.20191204.0853

吴言枫^{1, 2,},
王延杰^1,,
孙海江^1, ,,
刘培勋^1,

1.
中国科学院长春光学精密机械与物理研究所, 吉林长春 130033
2.
中国科学院大学材料科学与光电技术学院, 北京 100049

基金项目:

国家自然科学基金青年基金项目 61602432

详细信息

作者简介:
吴言枫(1992-), 男, 吉林长春人, 博士研究生, 2015年于吉林大学获得学士学位, 主要从事数字图像处理、计算机视觉方面的研究。E-mail:wuyfciomp@yahoo.com

王延杰(1963—)，男，吉林长春人，研究员，博士生导师，主要从事实时图像处理、电视跟踪和自动目标识别技术的研究。E-mail:wangyj@ciomp.ac.cn

孙海江(1981-), 男, 吉林辉南人, 副研究员, 主要从事目标识别与跟踪技术及高清视频图像增强显示方面的研究。E-mail:sunhaijiang@126.com

刘培勋(1986-), 男, 吉林榆树人, 助理研究员, 博士, 主要研究方向为数字图像处理及机器视觉。E-mail:liupx@ciomp.ac.cn

中图分类号: TP394.1;TH691.9
计量
- 文章访问数: 4313
- HTML全文浏览量: 1255
- PDF下载量: 288
- 被引次数: 0
出版历程
- 收稿日期: 2018-03-14
- 修回日期: 2018-04-04
- 刊出日期: 2019-08-01

LSS-target detection in complex sky backgrounds

WU Yan-feng^{1, 2
,},
WANG Yan-jie^1
,,
SUN Hai-jiang^{1
, ,},
LIU Pei-xun^1
,

1.
Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences, Changchun 130033, China
2.
University of Chinese Academy of Sciences, Beijing 100049, China

Funds:

the National Natural Science Foundation of China Youth Science Foundation Project 61602432

More Information

Corresponding author: SUN Hai-jiang, E-mail:sunhaijiang@126.com

摘要

摘要: 为了在复杂天空背景下检测出低空慢速小目标，本文研究了“低小慢”目标的视觉显著性区域特征，融合扫描线填充算法，提出了一种动态背景下“低小慢”目标自适应实时检测技术。首先，根据图像的亮度对比度获取显著性图。接着，使用形态学梯度提取显著性特征，通过三帧差分算法得到种子点。然后，使用扫描线填充算法进行生长，结合提出的自适应双高斯算法分割出前景。最后，根据候选目标的面积占比变化、质心距离变化、宽高比差异剔除虚假目标，完成检测。为了验证算法的有效性，本文选取了7组复杂天空背景的视频序列进行测试，并与其他优秀检测算法进行了对比。结果表明，本文提出的算法对运动目标检测的平均运行时间为0.040 9 s，平均检测准确率为89.97%，相比于其他算法的平均运算时间减少了0.35 s，检测的平均准确率提高了24.5%。算法在复杂背景下具有较好的稳定性和较强的鲁棒性。
- 计算机视觉 /
- 视觉显著性 /
- 扫描线填充 /
- 曲线拟合 /
- 自适应阈值分割
Abstract: In order to detect LSS(Low, Small and Slow) targets in complex sky backgrounds, we study the visual salient region characteristics of the LSS target and scan line filling algorithm and propose an adaptive real-time detection technology for LSS targets in dynamic complex backgrounds. Firstly, a saliency map is obtained based on the Luminance Contrast(LC) of the image. Secondly, the morphological gradient is used to extract the saliency feature and the seed points of the scan line filling algorithm are obtained by the three frame difference algorithm. Then, the scan line filling algorithm is used to grow the image and the foreground is segmented using the proposed adaptive double Gauss threshold segmentation algorithm. Finally, according to the change of the object's area of occupation, the center distance and the aspect ratio of the candidate target, the false targets are eliminated and detection is completed. In order to verify the effectiveness of the algorithm, 7 test groups of complex sky background video sequences are selected and compared with other excellent detection algorithms. The results show that the running time of the proposed algorithm for moving object detection is 0.040 9 s and the accuracy rate is 89.97%. When compared with other algorithms, the average running time is reduced by 0.35 s, and the average accuracy of detection is enhanced by 24.5%. The algorithm has good stability and is robust in target detection in complex backgrounds.
- computer vision /
- visual saliency /
- scan line filling /
- curve fitting /
- adaptive threshold segmentation

HTML全文

图 1 “低小慢”目标的特征分析图像

Figure 1. Feature analysis of the LSS-target image

下载: 全尺寸图片幻灯片

图 2 视觉显著性结果

Figure 2. Visual saliency results

下载: 全尺寸图片幻灯片

图 3 梯度差分算法原理图

Figure 3. Schematic diagram of gradient difference algorithm

下载: 全尺寸图片幻灯片

图 4 梯度差分结果

Figure 4. Gradient difference results

下载: 全尺寸图片幻灯片

图 5 扫描线填充算法流程图

Figure 5. Flow chart of scan line filling algorithm

下载: 全尺寸图片幻灯片

图 6 双高斯函数拟合直方图结果

Figure 6. Double Gauss function fitting histogram results

下载: 全尺寸图片幻灯片

图 7 自适应阈值分割结果

Figure 7. Adaptive threshold segmentation results

下载: 全尺寸图片幻灯片

图 8 对比实验结果

Figure 8. Comparison of experimental results for different algorithms

下载: 全尺寸图片幻灯片

图 9 本文算法检测结果

Figure 9. Detection results by proposed algorithm

下载: 全尺寸图片幻灯片

表 1 剔除虚警流程

Table 1. Eliminating false alarm process

input：候选目标集合R
output：目标集合T
1.t_i={r_i}, T={ϕ}	Δ初始化
2. For i=1, r_i∈R
3. if
4.	Δ更新目标集合
5. else
6. T_exist=T	Δ剔除虚警
7. End for

下载: 导出CSV

表 2 实验中的测试视频

Table 2. Test sequences in our experiments

视频	帧数	SCR
Video 1	76	＜1
Video 2	69	1~1.5
Video 3	249	1.5~2
Video 4	90	2~3
Video 5	22	＞3
Video 6	64	＞3
Video 7	101	＞3

下载: 导出CSV

表 3 检测准确率P_d及虚警率P_fa

Table 3. Detection accuracy and false alarm rate (%)

视频	准确率			虚警率
视频	Vibe	PBAS	Ours	Vibe	PBAS	Ours
Video 1	26.3	65. 8	81.6	77.2	35.8	5.1
Video 2	30.4	72.5	87.0	82.4	76.5	4.6
Video 3	32.1	80.3	84.3	65.3	11.2	0.9
Video 4	23.3	80	87.8	86.5	3.5	0.3
Video 5	45. 5	90.9	100	40.7	4.6	0
Video 6	62.5	78.1	100	33.5	4.8	0
Video 7	59.4	84.2	89.1	20.9	6.5	6.2

下载: 导出CSV

表 4 算法的时间复杂度

Table 4. Average time consumption and total time consumption of the proposed algorithm

视频	帧数	平均耗时(s/frame)			总耗时/s
视频	帧数	Vibe	PBAS	Ours	Vibe	PBAS	Ours
Video 1	76	0.362	0.396	0.042	27.512	30.096	3.05
Video 2	69	0.346	0.389	0.048	24.081	26.841	3.31
Video 3	249	0.372	0.412	0.047	96.682	102.588	11.61
Video 4	90	0.329	0.386	0.046	29.61	34.74	4.102
Video 5	22	0.349	0.391	0.045	7.678	8.602	0.988
Video 6	64	0.354	0.397	0.043	22.656	25.408	2.764
Video 7	101	0.371	0.403	0.047	37.471	40.703	4.770

下载: 导出CSV

参考文献(35)

[1]	HU J, YU Y, LIU F. Small and dim target detection by background estimation[J]. Infrared Physics & Technology, 2015, 73:141-148. http://d.old.wanfangdata.com.cn/Periodical/jgyhw201710023
[2]	YANG W P, LU X P, LI J CH, et al.. Fast algorithm of infrared small target detection in jitter background[J]. Proceedings of SPIE, 2015, 9476:947614.
[3]	DENG H, LIU J G, CHEN ZH. Infrared small target detection based on modified local entropy and EMD[J]. Chinese Optics Letters, 2010, 8(1):24-28. doi: 10.3788/COL20100801.0024
[4]	KIM S, YANG Y, LEE J, et al.. Robust scale invariant small target detection using the Laplacian scale-space theory[J]. Proceedings of SPIE, 2008, 6969:696911. doi: 10.1117/12.786354
[5]	姜志国, 韩冬兵, 袁天云, 等.基于全自动控制显微镜的自动聚焦算法研究[J].中国图象图形学报, 2004, 9(4):396-401. doi: 10.3969/j.issn.1006-8961.2004.04.003 JIANG ZH G, HAN D B, YUAN T Y, et al.. Study on auto focusing algorithm for automatic microscope[J]. Journal of Image and Graphics, 2004, 9(4):396-401.(in Chinese) doi: 10.3969/j.issn.1006-8961.2004.04.003
[6]	ABDELKAWY E, MCGAUGHY D. Small infrared target detection using two-dimensional fast orthogonal search(2D-FOS)[J]. Proceedings of SPIE, 2003, 5094:179-185. doi: 10.1117/12.487378
[7]	解婷, 陈忠, 马荣毅.一种基于PGF、BEMD和局部逆熵的新型红外小目标检测方法[J].红外与毫米波学报, 2017, 36(1):92-101. http://d.old.wanfangdata.com.cn/Periodical/hwyhmb201701017 XIE T, CHEN ZH, MA R Y. A novel method for infrared small target detection based on PGF, BEMD and LIE[J]. Journal of Infrared and Millimeter Waves, 2017, 36(1):92-101.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/hwyhmb201701017
[8]	BAI X ZH, ZHOU F G. Analysis of new top-hat transformation and the application for infrared dim small target detection[J]. Pattern Recognition, 2010, 43(6):2145-2156. doi: 10.1016/j.patcog.2009.12.023
[9]	WANG P, TIAN J W, GAO CH Q. Infrared small target detection using directional highpass filters based on LS-SVM[J]. Electronics Letters, 2009, 45(3):156-158. doi: 10.1049/el:20092206
[10]	REED I S, GAGLIARDI R M, STOTTS L B. Optical moving target detection with 3-D matched filtering[J]. IEEE Transactions on Aerospace and Electronic Systems, 2002, 24(4):327-336.
[11]	GUARIGLIA E. Entropy and fractal antennas[J]. Entropy, 2016, 18(3):84. doi: 10.3390/e18030084
[12]	SHEN X M, DENG L. Game theory approach to discrete H∞filter design[J]. IEEE Transactions on Signal Processing, 1997, 45(4):1092-1095. doi: 10.1109/78.564201
[13]	刘伟宁.基于小波域扩散滤波的弱小目标检测[J].中国光学, 2011, 4(5):503-508. doi: 10.3969/j.issn.2095-1531.2011.05.015 LIU W N. Dim target detection based on wavelet field diffusion filter[J]. Chinese Optics, 2011, 4(5):503-508.(in Chinese) doi: 10.3969/j.issn.2095-1531.2011.05.015
[14]	刘希佳, 陈宇, 王文生, 等.小目标识别的小波阈值去噪方法[J].中国光学, 2012, 5(3):248-256. doi: 10.3969/j.issn.2095-1531.2012.03.009 LIU X J, CHEN Y, WANG W SH, et al.. Denoising algorithm of wavelet threshold for small target detection[J]. Chinese Optics, 2012, 5(3):248-256.(in Chinese) doi: 10.3969/j.issn.2095-1531.2012.03.009
[15]	CASELLES V, KIMMEL R, SAPIRO G. Geodesic active contours[J]. International Journal of Computer Vision, 1997, 22(1):61-79. http://d.old.wanfangdata.com.cn/Periodical/jsjyjyfz200302024
[16]	LIU ZH J, CHEN CH Y, SHEN X B, et al.. Detection of small objects in image data based on the nonlinear principal component analysis neural network[J]. Optical Engineering, 2005, 44(9):9093604. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=d66c10351f8cf89b2fb5386e8beca31f
[17]	郭文佳, 朱晓辉, 王向军.野外图像监测中地平线附近的弱小目标的实时检测[C].光电技术与系统文选——中国光学学会光电技术专业委员会成立二十周年暨第十一届全国光电技术与系统学术会议论文集, 中国光学学会, 2005. GUO W J, ZHU X H, WANG X J. Real-time detection of weak and small targets near the horizon in field image monitoring[C]. Optoelectronic Technology of China Institute of Optics and Optoelectronic Technology Specialized Committee Establishment 20th Anniversary and National Photoelectric Technology And System Academic Conference, Chinese Optical Society, 2005.(in Chinese)
[18]	胡谋法, 王书宏, 李超, 等.空时域联合去相关检测可见光背景下的运动小目标[C].光电技术与系统文选——中国光学学会光电技术专业委员会成立二十周年暨第十一届全国光电技术与系统学术会议论文集, 中国光学学会, 2005. HU M F, WANG SH H, LI CH, et al.. Detection of small moving targets in the visible light background by the joint spatio-temporal correlation detection[C]. Optoelectronic Technology of China Institute of Optics and Optoelectronic Technology Specialized Committee Establishment 20th Anniversary and National Photoelectric Technology and System Academic Conference, Chinese Optical Society, 2005.(in Chinese)
[19]	洪普, 王群, 阮建斌, 等.复杂背景红外弱点目标检测[C].中国光学学会2010年光学大会论文集, 中国光学学会, 2010. HONG P, WANG Q, RUAN J B, et al.. Infrared weak target detection in complex background[C]. A Collection of the 2010 Optical Conference of the Chinese Academy of Optics, hinese Optical Society, 2010.(in Chinese)
[20]	姚成乾, 陈伟.基于改进粒子算法的红外弱小目标检测研究[J].激光与光电子学进展, 2017, 54(11):111101. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=jgygdzxjz201711017 YAO CH Q, CHEN W. Infrared dim target detection based on improved particle swarm optimization algorithm[J]. Laser & Optoelectronics Progress, 2017, 54(11):111101.(in Chinese) http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=jgygdzxjz201711017
[21]	刘靳, 姬红兵.基于移动式加权管道滤波的红外弱小目标检测[J].西安电子科技大学学报(自然科学版), 2007, 34(5):743-747. doi: 10.3969/j.issn.1001-2400.2007.05.013 LIU J, JI H B. Detection method for small targets in the IR image based on the variable weighted pipeline filter[J]. Journal of Xidian University, 2007, 34(5):743-747.(in Chinese) doi: 10.3969/j.issn.1001-2400.2007.05.013
[22]	黄宇, 张晓芳, 俞信.光子成像静止点目标的管道滤波探测方法[J].中国光学, 2013, 6(1):73-79. http://www.chineseoptics.net.cn/CN/abstract/abstract8893.shtml HUANG Y, ZHANG X F, YU X. Pipeline filtering detection of stationary point targets in photon images[J]. Chinese Optics, 2013, 6(1):73-79.(in Chinese) http://www.chineseoptics.net.cn/CN/abstract/abstract8893.shtml
[23]	NIE H SH, HUANG ZH J, DIAO J T, et al.. A wiener filter based infrared small target detecting and tracking method[C]. Proceedings of 2010 International Conference on Intelligent System Design and Engineering Application, IEEE, 2010: 184-187.
[24]	ZHAI Y, SHAH M. Visual attention detection in video sequences using spatiotemporal cues[C]. Proceedings of the 14th ACM International Conference on Multimedia, ACM, 2006: 815-824.
[25]	王晓阳, 彭真明, 张萍, 等.局部对比度结合区域显著性红外弱小目标检测[J].强激光与粒子束, 2015, 27(9):32-38. http://d.old.wanfangdata.com.cn/Periodical/qjgylzs201509006 WANG X Y, PENG ZH M, ZHANG P, et al.. Infrared small dim target detection based on local contrast combined with region saliency[J]. High Power Laser and Particle Beams, 2015, 27(9):32-38.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/qjgylzs201509006
[26]	毋亚北, 王卫华, 吴巨红, 等.基于形态学梯度的红外目标检测[J].光电工程, 2012, 39(9):81-85. http://d.old.wanfangdata.com.cn/Periodical/gdgc201209015 WU Y B, WANG W H, WU J H, et al..IR target detection based on morphological gradient[J]. Opto-Electronic Engineering, 2012, 39(9):81-85.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/gdgc201209015
[27]	ZHANG Y Z, WANG X Y, QU B. Three-frame difference algorithm research based on mathematical morphology[J]. Procedia Engineering, 2012, 29:2705-2709. doi: 10.1016/j.proeng.2012.01.376
[28]	GONZALEZ R C, WOODS R E. Digital Image Processing[M]. RUAN Q Q, RUAN Y ZH, trans. 2nd ed. Beijing: Publishing House of Electronics Industry, 2007.
[29]	肖丽君, 于哲舟, 周栩, 等.基于对称差分算法的视频运动目标分割[J].吉林大学学报(理学版), 2008, 46(4):691-696. doi: 10.3321/j.issn:1671-5489.2008.04.025 XIAO L J, YU ZH ZH, ZHOU X, et al.. Moving object segmentation based on symmetrical differencing algorithm in video sequences[J]. Journal of Jilin University(Science Edition), 2008, 46(4):691-696.(in Chinese) doi: 10.3321/j.issn:1671-5489.2008.04.025
[30]	范宏深, 倪国强, 冯煜芳.复杂背景可见光图像中弱小目标探测的新算法[J].光电工程, 2004, 31(6):48-51. doi: 10.3969/j.issn.1003-501X.2004.06.014 FAN H SH, NI G Q, FENG Y F. A new algorithm for small and dim target detection of visible image under heavy clutters[J]. Opto-Electronic Engineering, 2004, 31(6):48-51.(in Chinese) doi: 10.3969/j.issn.1003-501X.2004.06.014
[31]	WU Y F, SUN H J, LIU P X. A novel fast detection method of infrared LSS-target in complex urban background[J]. International Journal of Wavelets, Multiresolution and Information Processing, 2018, 16(1):1850008. doi: 10.1142/S021969131850008X
[32]	BARNICH O, VAN DROOGENBROECK M. ViBE: A powerful random technique to estimate the background in video sequences[C]. Proceedings of 2009 IEEE International Conference on Acoustics, Speech and Signal Processing, IEEE, 2009: 945-948.
[33]	HOFMANN M, TIEFENBACHER P, RIGOLL G. Background segmentation with feedback: the pixel-based adaptive segmenter[C]. Proceedings of 2012 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, IEEE, 2012: 38-43.
[34]	王敏敏, 孙胜利.并行压缩成像系统的压缩域小目标检测[J].光学精密工程, 2016, 34(10):2549-2556. http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201610025 WANG M M, SUN SH L. Small target detection in compressed domain of parallel compression imaging system[J]. Opt. Precision Eng., 2016, 34(10):2549-2556.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201610025
[35]	李正周, 曹雷, 邵万兴, 等.基于空时混沌分析的海面小弱目标检测[J].光学精密工程, 2018, 26(1):193-199. http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201801023 LI ZH ZH, CAO L, SHAO W X, et al.. Detection of small and weak targets in sea surface based on spatio temporal chaos analysis[J]. Opt. Precision Eng., 2018, 26(1):193-199.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201801023