复杂动态背景下基于群稀疏的运动目标检测方法

doi:10.12263/DZXB.20200816

电子学报 ›› 2021, Vol. 49 ›› Issue (12): 2330-2338.DOI: 10.12263/DZXB.20200816

复杂动态背景下基于群稀疏的运动目标检测方法

王洪雁¹^,², 张海坤², 罗宇华³, 汪祖民²

^1.浙江理工大学信息学院，浙江杭州 310018
^2.大连大学信息工程学院，辽宁大连 116622
^3.上海精密计量测试研究所，上海 201109

收稿日期:2020-07-31 修回日期:2021-06-15 出版日期:2021-12-25
- 作者简介:
- 王洪雁男,1979年5月生于河南南阳.2011年毕业于西安电子科技大学获信号与信息处理专业博士学位,现为浙江理工大学信息学院特聘教授、硕士生导师.主要研究方向为:阵列信号处理、机器视觉、深度学习等.E-mail:gglongs@163.com
  张海坤男，1995年4月生于安徽阜阳.现为大连大学信息学院控制科学与工程专业硕士研究生，主要研究方向为计算机视觉.E-mail:707274821@qq.com
  罗宇华男,1979年生于江西吉安,上海精密计量测试研究所高级工程师, 研究方向为图像处理、元器件可靠性.E-mail:luoyuhua@163.com
  汪祖民男,1975年生于河南省信阳，大连大学信息工程学院教授、硕士生导师,研究方向为物联网、智慧城市.E-mail:wangzumin@dlu.edu.cn
- 基金资助:
- 国家自然科学基金 (61301258); 浙江省自然科学基金重点项目 (LZ21F010002); 中国博士后科学基金资助项目 (2016M590218)

Moving Object Detection Method Based on Group Sparseness Under Complex Dynamic Background

WANG Hong-yan¹^,², ZHANG Hai-kun², LUO Yu-hua³, WANG Zu-min²

^1.School of Information Science and Technology, Zhejiang Sci-Tech University, Hangzhou, Zhejiang 310018, China
^2.College of Information Engineering, Dalian University, Dalian, Liaoning 116622, China
^3.Shanghai Precision Metrology & Test Research Institute, Shanghai 201109, China

Received:2020-07-31 Revised:2021-06-15 Online:2021-12-25 Published:2021-12-25
- Supported by:
- National Natural Science Foundation of China (61301258); Key Program of National Natural Science Foundation of Zhejiang Province, China (LZ21F010002); Supported by China Postdoctoral Science Foundation (2016M590218)

摘要/Abstract

摘要：

为提高复杂动态背景下运动目标检测精度，基于低秩及稀疏分解理论，本文提出一种基于群稀疏的运动目标检测方法. 所提方法将观测视频分解为低秩静态背景，群稀疏前景及动态背景三部分. 所提方法首先使用伽马范数近乎无偏近似矩阵秩函数，以解决核范数过度惩罚较大奇异值导致所得最小化问题无法获得最优解进而降低检测性能的问题；其次，为利用前景目标边界先验信息以提升运动目标检测性能，每一帧使用过分割算法生成同性区域以定义群稀疏范数并用于约束前景矩阵；再次，为避免运动目标同时出现在稀疏前景和动态背景中，引入非相干项以提升二者可分性；最后，本文利用交替方向乘子方法（Alternating Direction Method of Multipliers，ADMM）求解所得非凸目标函数. 实验结果表明，与现有主流运动目标检测算法相比，复杂动态背景下本文所提方法可较好抑制动态背景从而显著提高复杂运动背景下运动目标检测精度.

关键词: 运动目标检测, 动态背景, 低秩, 群稀疏, 超像素

Abstract:

To improve the accuracy of moving object detection under complex dynamic background, based on the theory of low-rank and sparse decomposition, a group sparse based moving object detection method is developed. The proposed method decomposes the observed video into a low-rank static background, a group sparse foreground and a dynamic background. Regarding the problem that the nuclear norm over-penalizing large singular values leads to the optimal solution of the obtained minimization problem cannot be obtained and then the detection performance is decreased，the gamma norm is introduced to acquire almost unbiased approximation of rank function. In order to utilize the object boundary prior to enhance the moving target detection performance, each frame is over-segmented into homogeneous regions which are taken to define the group sparse norm to constrain the foreground matrix. Moreover, to prevent the moving object from appearing in the sparse foreground and dynamic background simultaneously, the incoherence term is introduced to enhance the separability of them. Finally, the obtained non-convex objective function can be solved using the alternating direction multiplier method（ADMM）. The experimental results show that, compared with the state-of-the-art moving target detection algorithms, the developed method can suppress the dynamic background considerably and then improve the accuracy of moving object detection significantly under complex dynamic background.

Key words: moving object detection, dynamic background, low-rank, group sparse, superpixel

中图分类号:

TP391

王洪雁, 张海坤, 罗宇华, 等. 复杂动态背景下基于群稀疏的运动目标检测方法[J]. 电子学报, 2021, 49(12): 2330-2338.

Hong-yan WANG, Hai-kun ZHANG, Yu-hua LUO, et al. Moving Object Detection Method Based on Group Sparseness Under Complex Dynamic Background[J]. Acta Electronica Sinica, 2021, 49(12): 2330-2338.

图/表 6

参考文献 30

1	YangY C, LoquercioA, ScaramuzzaD, et al. Unsupervised moving object detection via contextual information separation[A]. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) [C]. Long Beach, CA, USA: IEEE, 2019. 879 - 888.
2	SaxenaT, TripathiV, ChandolaA, et al. Robust moving object detection and tracking framework using linear phase FIR filter[A]. International Conference on Emerging Technologies in Computer Engineering[C]. Singapore: Springer, 2019. 34 - 44.
3	B S, TomA J, GeorgeS N. Simultaneous denoising and moving object detection using low rank approximation[J]. Future Generation Computer Systems, 2019, 90: 198 - 210.
4	CandèsE J, LiX D, MaY, et al. Robust principal component analysis[J]. Journal of the ACM, 2011, 58(3): 1 - 37.
5	XUEYa-wen, GUOXiao-jie, CAOXiao-chun. Motion saliency detection using low-rank and sparse decomposition[A]. 2012 IEEE International Conference on Acoustics, Speech and Signal Processing[C]. Kyoto, Japan: IEEE, 2012. 1485 - 1488.
6	ZHOUTian-yi, TAODa-cheng. Godec: Randomized low-rank & sparse matrix decomposition in noisy case[A]. Proceedings of the 28th International Conference on Machine Learning[C]. Sydney: ICML,2011. 33 - 40.
7	WANGNai-yan, YAOTian-sheng, WANGJing-dong, et al. A probabilistic approach to robust matrix factorization. European Conference on Computer Vision[C]. Berlin, Germany: Springer, 2012. 126 - 139.
8	ZhouX W, YangC, YuW C. Moving object detection by detecting contiguous outliers in the low-rank representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(3): 597 - 610.
9	ZHAOQian, MENGDe-yu, XUZong-ben, et al. Robust principal component analysis with complex noise[A]. 31st International Conference on Machine Learning[C]. Beijing, China: Microtome,2014. 55 - 63.
10	周伟, 孙玉宝, 刘青山, 等. 运动目标检测的l₀群稀疏RPCA模型及其算法[J]. 电子学报, 2016, 44(3): 627 - 632.
	ZhouW, SunY B, LiuQ S, et al. l₀ group sparse RPCA model and algorithm for moving object detection[J]. Acta Electronica Sinica, 2016, 44(3): 627 - 632. (in Chinese)
11	GaoZ, CheongL F, WangY X. Block-sparse RPCA for salient motion detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014, 36(10): 1975 - 1987.
12	AchantaR, ShajiA, SmithK, et al. SLIC Superpixels[R]. Lausanne, Swiss Confederation: EPFL, 2010.
13	ZhangH M, YangJ, XieJ C, et al. Weighted sparse coding regularized nonconvex matrix regression for robust face recognition[J]. Information Sciences, 2017, 394/395: 1 - 17.
14	WANGShu-sen, LIUDe-hua, ZHANGZhi-hua. Nonconvex relaxation approaches to robust matrix recovery[A]. Twenty-Third International Joint Conference on Artificial Intelligence[C].Beijing, China: AAAI Press, 2013. 1764 - 1770.
15	CaoW F, SunJ, XuZ B. Fast image deconvolution using closed-form thresholding formulas of Lq(Q=12, 23) regularization[J]. Journal of Visual Communication and Image Representation, 2013, 24(1): 31 - 41.
16	ShiJ B, MalikJ. Normalized cuts and image segmentation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(8): 888 - 905.
17	LevinshteinA, StereA, KutulakosK N, et al. TurboPixels: Fast superpixels using geometric flows[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(12): 2290 - 2297.
18	AchantaR, ShajiA, SmithK, et al. SLIC superpixels compared to state-of-the-art superpixel methods[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2012, 34(11): 2274 - 2282.
19	SunY B, LiuQ S, TangJ H, et al. Learning discriminative dictionary for group sparse representation[J]. IEEE Transactions on Image Processing, 2014, 23(9): 3816 - 3828.
20	LIUBo, YUANXiao-tong, YUYang, et al. Decentralized robust subspace clustering[A]. Thirtieth AAAI Conference on Artificial Intelligence[C]. Phoenix, Arizona USA: AAAI Press,2016. 3539 - 3545.
21	王佰玲, 田志宏, 张永铮. 奇异值分解算法优化[J]. 电子学报, 2010, 38(10): 2234 - 2239.
	WangB L, TianZ H, ZhangY Z. Optimization of singular vector decomposition algorithm[J]. Acta Electronica Sinica, 2010, 38(10): 2234 - 2239. (in Chinese)
22	LiuG C, LinZ C, YanS C, et al. Robust recovery of subspace structures by low-rank representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(1): 171 - 184.
23	LINZhou-chen, CHENMin-ming, et al. The augmented lagrange multiplier method for exact recovery of corrupted low-rank matrices[EB/OL]. , 2010.
24	LiL Y, HuangW M, GuIreneY H, et al. Statistical modeling of complex backgrounds for foreground object detection[J]. IEEE Transactions on Image Processing, 2004, 13(11): 1459 - 1472.
25	MahadevanV, VasconcelosN. Spatiotemporal saliency in dynamic scenes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2010, 32(1): 171 - 177.
26	ToyamaK, KrummJ, BrumittB, et al. Wallflower: Principles and practice of background maintenance[A]. Proceedings of The Seventh IEEE International Conference on Computer Vision[C]. Kerkyra, Greece: IEEE, 1999. 255 - 261.
27	WangYi, JodoinP M, PorikliF, et al. CDnet 2014: An expanded change detection benchmark dataset[A]. 2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops[C]. Columbus, USA: IEEE, 2014. 393 - 400.
28	YongH W, MengD Y, ZuoW M, et al. Robust online matrix factorization for dynamic background subtraction[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2018, 40(7): 1726 - 1740.
29	LiuX, ZhaoG Y, YaoJ W, et al. Background subtraction based on low-rank and structured sparse decomposition[J]. IEEE Transactions on Image Processing, 2015, 24(8): 2502 - 2514.
30	YangH H, QuS R. Real-time vehicle detection and counting in complex traffic scenes using background subtraction model with low-rank decomposition[J]. IET Intelligent Transport Systems, 2018, 12(1): 75 - 85.

视频序列	基本信息	挑战因素
Airport	室内场景，176*144分辨率，RGB图像	遮挡
SwitchLight	室内场景，160*128分辨率，RGB图像	光照变化
WaterSurface	室外场景，160*128分辨率，RGB图像	动态背景，徘徊目标
Rain	室外场景，308*228分辨率，灰度图像	动态背景
WavingTrees	室外场景，160*120分辨率，RGB图像	动态背景
Boats	室外场景，320*240分辨率，RGB图像	动态背景
Bottle	室外场景，304*224分辨率，灰度图像	动态背景
Fountain	室外场景，432*288分辨率，RGB图像	动态背景
Bootstrap	室内场景，160*120分辨率，RGB图像	遮挡，动态背景
Camouflage	室内场景，160*120分辨率，RGB图像	动态背景
ForegroundAperture	室内场景，160*120分辨率，RGB图像	动态背景

视频序列	基本信息	挑战因素
Airport	室内场景，176*144分辨率，RGB图像	遮挡
SwitchLight	室内场景，160*128分辨率，RGB图像	光照变化
WaterSurface	室外场景，160*128分辨率，RGB图像	动态背景，徘徊目标
Rain	室外场景，308*228分辨率，灰度图像	动态背景
WavingTrees	室外场景，160*120分辨率，RGB图像	动态背景
Boats	室外场景，320*240分辨率，RGB图像	动态背景
Bottle	室外场景，304*224分辨率，灰度图像	动态背景
Fountain	室外场景，432*288分辨率，RGB图像	动态背景
Bootstrap	室内场景，160*120分辨率，RGB图像	遮挡，动态背景
Camouflage	室内场景，160*120分辨率，RGB图像	动态背景
ForegroundAperture	室内场景，160*120分辨率，RGB图像	动态背景

评价指标	RPCA	GoDec	PRMF	DEC	MoG	OMoGMF	LSD	本文
Precision	0.5840	0.4655	0.6164	0.6574	0.6103	0.6836	0.6244	0.9139
Recall	0.5873	0.6052	0.5497	0.6030	0.6014	0.5265	0.5542	0.7744
F-measure	0.4932	0.4516	0.5108	0.5718	0.5427	0.5402	0.5446	0.8293

评价指标	RPCA	GoDec	PRMF	DEC	MoG	OMoGMF	LSD	本文
Precision	0.5840	0.4655	0.6164	0.6574	0.6103	0.6836	0.6244	0.9139
Recall	0.5873	0.6052	0.5497	0.6030	0.6014	0.5265	0.5542	0.7744
F-measure	0.4932	0.4516	0.5108	0.5718	0.5427	0.5402	0.5446	0.8293

[1]	李敏, 栗泽昊, 张卓, 张学武. 融合局部低秩先验与Bloch流形约束的磁共振指纹重建算法[J]. 电子学报, 2023, 51(2): 372-384.
[2]	金天虎, 陶砚蕴, 李佐勇. 基于超像素图像分割的暗通道先验去雾改进算法[J]. 电子学报, 2023, 51(1): 146-159.
[3]	陈震, 倪晶晶, 张聪炫, 葛利跃, 王梓歌. 基于金字塔块匹配的双目场景流估计[J]. 电子学报, 2022, 50(9): 2164-2171.
[4]	周伯阳, 郭志民, 王延松, 阮伟, 吴春明, 周宁, 张伟, 程国振. 基于多尺度低秩模型的电力无线接入网异常流量检测方法[J]. 电子学报, 2020, 48(8): 1552-1557.
[5]	陈震, 马龙, 张聪炫, 黎明, 吴俊劼, 江少锋. 基于语义分割的双目场景流估计[J]. 电子学报, 2020, 48(4): 631-636.
[6]	欧先锋, 晏鹏程, 王汉谱, 涂兵, 何伟, 张国云, 徐智. 基于深度帧差卷积神经网络的运动目标检测方法研究[J]. 电子学报, 2020, 48(12): 2384-2393.
[7]	葛利跃, 张聪炫, 陈震, 黎明, 陈昊. 相互结构引导滤波TV-L1变分光流估计[J]. 电子学报, 2019, 47(3): 707-713.
[8]	范虹, 张程程, 侯存存, 朱艳春, 姚若侠. 结合双树复小波变换和改进密度峰值快速搜索聚类的乳腺MR图像分割[J]. 电子学报, 2019, 47(10): 2149-2157.
[9]	侯小刚, 赵海英, 马严. 基于超像素多特征融合的快速图像分割算法[J]. 电子学报, 2019, 47(10): 2126-2133.
[10]	陈允杰, 马辰阳, 孙乐, 詹天明. 基于边缘修正的高光谱图像超像素空谱核分类方法[J]. 电子学报, 2019, 47(1): 73-81.
[11]	李磊, 董卓莉, 张德贤. 基于自适应区域限制FCM的图像分割方法[J]. 电子学报, 2018, 46(6): 1312-1318.
[12]	张长伦, 余沾, 王恒友, 何强. 基于广义低秩矩阵分解的分离字典训练及其快速重建算法[J]. 电子学报, 2018, 46(10): 2400-2409.
[13]	蒋明峰, 陆雨, 朱志军, 徐文龙, 汪亚明. 基于多尺度低秩模型的心脏磁共振成像方法研究[J]. 电子学报, 2017, 45(9): 2218-2224.
[14]	常侃, 张智勇, 陈诚, 覃团发. 采用低秩与加权稀疏分解的视频前景检测算法[J]. 电子学报, 2017, 45(9): 2272-2280.
[15]	宋云, 李雪玉, 沈燕飞, 杨高波. 基于非局部相似块低秩的压缩感知图像重建算法[J]. 电子学报, 2017, 45(3): 695-703.

复杂动态背景下基于群稀疏的运动目标检测方法

Moving Object Detection Method Based on Group Sparseness Under Complex Dynamic Background

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 6

参考文献 30

相关文章 15

编辑推荐

Metrics