结合阴影抑制的混合高斯模型改进算法

J4 ›› 2016, Vol. 38 ›› Issue (03): 556-561.

结合阴影抑制的混合高斯模型改进算法

李博川1，丁轲2

（1.合肥工业大学电气与自动化工程学院，安徽合肥 230009;2.合肥华耀电子工业有限公司，安徽合肥 230088）

收稿日期:2015-03-09 修回日期:2015-05-22 出版日期:2016-03-25 发布日期:2016-03-25
基金资助:
国家国际科技合作项目（2011FA10440）

An improved algorithm of Gaussian mixture
model combined with shadow suppression

LI Bochuan1,DING Ke2

（1.School of Electrical Engineering and Automation,Hefei University of Technology,Hefei 230009;
2.ECU Electronics Industrial Co.,LTD.,Hefei 230088,China）

Received:2015-03-09 Revised:2015-05-22 Online:2016-03-25 Published:2016-03-25

摘要/Abstract

摘要：

混合高斯模型背景法作为运动目标检测的一种经典方法，已经广泛应用于智能视频监控系统中。但是，传统的混合高斯模型背景法容易将阴影误检测为运动目标的一部分。因此，针对该方法在区分阴影和运动目标方面的不足，提出了一种将混合高斯模型背景法和HSV空间阴影抑制相结合的运动目标检测算法。这种改进算法首先将颜色空间转换到HSV空间，初步提取运动目标，然后再利用阴影的灰度值比背景中的灰度值小，而前景的灰度值比背景中灰度值大的特性，检测出运动目标中的阴影。实验结果表明，这种改进的算法明显提高了检测效果，有效抑制了阴影对运动目标检测的干扰，算法实时性也较好。

关键词: 运动目标检测, 混合高斯模型, HSV空间, 阴影抑制

Abstract:

As a classical method in moving target detection, the background subtraction of the Gaussian mixture model has been widely applied in intelligent video surveillance system. However, this classical method easily recognizes shadows as a part of a moving target. So in this paper, we present a moving target detection algorithm combining the Gaussian mixture model and the shadow suppression in HSV space to overcome the shortage in distinguishing shadows from a moving target. First, we use the three frames subtraction method to detect the changed area in the image. Then we transform RGB space into HSV space to detect moving targets preliminarily. Finally, we detect shadows from the moving target through the features that the gray value of shadows is smaller than that of background and the gray value of the foreground is larger than that of the background. Experimental results suggest that the improved algorithm can obviously improve the detection effect by suppressing the interference of shadows of moving targets and also has good realtime performance.

Key words: moving target detection;Gaussian mixture model;HSV space;shadow suppression

李博川1，丁轲2. 结合阴影抑制的混合高斯模型改进算法[J]. J4, 2016, 38(03): 556-561.

LI Bochuan1,DING Ke2. An improved algorithm of Gaussian mixture
model combined with shadow suppression [J]. J4, 2016, 38(03): 556-561.

参考文献［11］

［1］	Lin Lizhong,Liu Zhiguo,Zhang Yubin.Research on detection and tracking of moving target in intelligent video surveillance［C］∥Proc of International Conference on Computer Science and Electronics Engineering,2012:477481.
［2］	Stauffer C, Grimson W. Learning patterns of activity using realtime tracking ［J］.IEEE Transactions on Pattern Analysis and Machine Intelligence,2000,22（8）:747757.
［3］	Liu Guangyu,Liu Guodong.Method for moving objection detection based on mixture Gaussian models ［J］.Computer Engineering and Applications,2009,45（24）:180182.（in Chinese）
［4］	Han Yanxiang, Zhang Zhisheng. An efficient estimation method for intensity factor of illumination changes ［J］.Multimedia Tools and Applications,2014,72（3）:26192632.
［5］	Chen Zezhi,Ellis T.A selfadaptive Gaussian mixture model ［J］.Computer Vision and Image Understanding,2014,122:3546.
［6］	Li Quanming, Zhang Yunchu. Improvement on adaptive mixture Gaussian background model ［J］.Journal of Computer Applications,2007,27（8）:20142017.（in Chinese）
［7］	Zivkovic Z. Improved adaptive Gaussian mixture model for background subtraction［C］∥Proc of the 17th International Conference on Pattern Recognition,2004:2831.
［8］	Evangelio R H,Patzold M,Keller I,et al.Adaptively splitted GMM with feedback improvement for the task of background subtraction ［J］.IEEE Transactions on Information Forensics and Security,2014,9（5）:863874.
［9］	Liu Jing, Wang Ling.Improved algorithm of Gaussian mixture model for background subtraction ［J］.Computer Engineering and Applications,2010,46（13）:168170.（in Chinese）
［10］	Stauffer C,Grimson W.Adaptive background mixture models for realtime tracking［C］∥Proc of IEEE International Conference on Computer Vision and Pattern Recognition,1999:246252.
［11］	Zhang Li, Li Zhineng. Adaptive HSV color background modeling for realtime vehicle tracking with shadow detection in traffic surveillance ［J］.Journal of Image and Graphics,2003,8（7）:6064.（in Chinese）
	附中文参考文献：
［3］	刘光宇,刘国栋.基于混合高斯模型的运动目标检测方法［J］.计算机工程与应用,2009,45（24）:180182.
［6］	李全民，张运楚.自适应混合高斯背景模型的改进［J］.计算机应用,2007,27（8）:20142017.
［9］	刘静，王玲.混合高斯模型背景法的一种改进算法［J］.计算机工程与应用，2010,46（13）:168170.
［11］	张丽，李志能.基于阴影检测的HSV空间自适应背景模型的车辆追踪检测［J］.中国图像图形学报,2003,8（7）:6064.

[1]	张寅, 蔡旭阳, 许倩倩, 闫钧华, 苏恺, 张琨 . 融合时空上下文的复杂背景下多运动目标检测[J]. 计算机工程与科学, 2020, 42(12): 2186-2192.
[2]	马琴，张兴忠，李海芳，邓红霞. 基于谱残差和聚类法的运动目标检测研究[J]. 计算机工程与科学, 2018, 40(10): 1867-1873.
[3]	金静1,党建武1,王阳萍1,2,翟凤文1. 一种基于鲁棒局部纹理特征的背景差分方法[J]. 计算机工程与科学, 2017, 39(08): 1513-1519.
[4]	杨国亮，丰义琴，唐俊，谢乃俊. 基于低秩表示动态更新投影的在线运动目标检测[J]. 计算机工程与科学, 2016, 38(11): 2289-2296.
[5]	李伟生，李辉飞. 一种基于自适应混合模型的背景减除法[J]. 计算机工程与科学, 2016, 38(10): 2091-2100.
[6]	叶立仁，何盛鸿，赵连超. 复杂环境下的遗留物检测算法[J]. J4, 2015, 37(05): 986-992.
[7]	黄玉，殷苌茗，周书仁. 基于YCbCr的自适应混合高斯模型背景建模[J]. J4, 2015, 37(01): 152-156.
[8]	金圣韬1，孟朝晖1，刘伟2. 基于团块模型的行人阴影抑制算法[J]. J4, 2014, 36(11): 2203-2209.
[9]	徐蔚鸿，严金果. 基于改进的背景差分的运动目标实时检测算法[J]. J4, 2014, 36(07): 1352-1356.
[10]	刘军，梁久祯，柴志雷. 基于DM642的运动目标检测[J]. J4, 2013, 35(1): 107-112.
[11]	吴亮周东翔范才智蔡宣平. 基于直方图匹配的鬼影检测算法[J]. J4, 2008, 30(7): 57-60.
[12]	单勇[1,2] 王润生[2] 杨凡[1]. 基于多特征背景模型的运动目标检测算法[J]. J4, 2007, 29(8): 40-42.
[13]	倪崇嘉[1] 刘文奇[2] 张爱英[2]. 基于数学形态学的视频图像序列中的运动目标检测[J]. J4, 2006, 28(6): 69-70.