摘要
运动目标检测是智能视频监控中的关键问题.Vibe是一种典型的运动目标检测算法,但是这种方法存在对鬼影消除速度缓慢以及对全局光线变化的抗干扰性差等缺点.本文提出一种改进算法,改进Vibe的背景模型更新机制,引入三帧差法作为参考帧,提升了消除鬼影的速度和背景模型的稳定性.提出一种全局光线抗干扰策略,降低了全局光线对目标检测的干扰,并通过实验验证了本文算法的有效性和可行性.
Motion target detection is a key issue in intelligent video surveillance. Vibe is a typical algorithm of moving target detection,this method has the disadvantages of slow elimination of ‘ghosts' and poor anti-interference of global light changes. A new algorithm was proposed,which improved the background model update mechanism of Vibe,and the results of the three-frame difference method was used as a reference frame to improve the speed of ghosting elimination and the stability of the background model. A global light anti-jamming strategy was proposed to reduce the global light's interference to the target detection,and the effectiveness and feasibility of the proposed algorithm were verified by experiments.
引文
[1]郭玲.智能视频监控中运动目标检测的算法研究[D].广州:华南理工大学,2013.
[2]张肃,王文生,徐春云.低对比度环境下运动目标光学相关检测技术[J].仪器仪表学报,2013,34(2):319-325.
[3]Elgammal A,Harwood D,Davis L.Non-parametric model for background subtraction[C]//European conference on computer vision,Springer,Berlin,Heidelberg,2000:751-767.
[4]Lee D S.Effective Gaussian mixture learning for video background subtraction[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2005,27(5):827-32.
[5]Barnich O,Van Droogenbroeck M.Vi Be:a universal background subtraction algorithm for video sequences[J].IEEE Transactions on Image Processing,2011,20(6):1709-1724.
[6]Shoushtarian B,Bez H E.A practical adaptive approach for dynamic background subtraction using an invariant colour model and object tracking[J].Pattern Recognition Letters,2005,26(1):5-26.
[7]李晓娟,李佳田,何育枫.目标相似性度量的鬼影抑制方法[J].计算机应用研究,2014,31(3):926-928.
[8]Li Y,Chen W,Jiang R.The integration adjacent frame difference of improved Vi Be for foreground object detection[C]//International Conference on Wireless Communications,Networking and Mobile Computing,Wuhan,China,2011:1-4.
[9]魏洪涛,李瑾,吴细秀.融合帧差和Vi Be的运动目标检测算法[J].计算机应用研究,2017,34(5):1565-1567.
[10]闫硕,陈科山.基于双背景模型的鬼影抑制方法研究[J].计算机应用与软件,2016,33(5):162-165.
[11]陈亮,陈晓竹,胡正东.用于鬼影抑制的区域检测算法[J].中国计量学院学报,2015,26(1):115-122.
[12]Goyette N,Jodoin P M,Porikli F,et al.Changedetection.net:A new change detection benchmark dataset[C]//Computer Vision and Pattern Recognition Workshops,Providence,USA,2012:1-8.