一种快速消除鬼影的目标检测算法  

AN OBJECT DETECTION ALGORITHM WITH FAST GHOST ELIMINATION

作　　者：路霄汉 王志君[1] 梁利平[1] Lu Xiaohan;Wang Zhijun;Liang Liping(Institute of Microelectronics,Chinese Academy of Science,Beijing 100029,China;School of Electrical,Electronics and Communication Engineering,University of Chinese Academy of Sciences,Beijing 100029,China)

机构地区：[1]中国科学院微电子研究所,北京100029 [2]中国科学院大学电子电气与通信工程学院,北京100029

出　　处：《计算机应用与软件》2019年第10期101-106,129,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(61471354)

摘　　要：针对运动目标检测ViBe算法在执行过程中的鬼影以及背景抖动问题,提出一种可以快速消除鬼影的目标检测算法。提出背景值定向传播以及计算闪烁因子的优化算法,将背景值快速填充到周边鬼影区域的背景集合中,从而快速消除鬼影并消除动态背景和相机抖动的影响,避免使用复杂的特征点检测算法处理抖动问题。通过快速滤波与形态学操作可以去除噪点并提高前景的完整性。经过理论分析以及实验对比,结果表明,该算法能够快速消除鬼影,在相机抖动或动态背景的情况下有更好的目标检测效果,算法的运行速度相比部分主流算法更加快速高效。To eliminate the ghost and background jitter problem in moving object detection based on ViBe algorithm,we proposed an improved object detection algorithm with fast ghost elimination.An optimization algorithm for directional propagation of background values and calculation of scintillation factor was proposed,which could quickly fill the background values into the background set of surrounding ghost regions.Thus it could rapidly eliminate the ghost and remove the influence of camera jittering and dynamic background.The complex feature point detection algorithm was avoided to deal with the jitter problem.We used fast filtering and morphological operations to remove noise and improve the integrity of the foreground.The theoretical analysis and the experimental results show that the improved ViBe algorithm can quickly eliminate ghosts and has better recognition results in the case of camera jitter or dynamic background.The proposed algorithm runs faster and more efficiently than some main popular algorithms.

关 键 词：运动目标检测 ViBe 鬼影消除 抖动处理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]