基于视频的运动目标检测跟踪模型的应用研究  

Research on Application of Video-based Moving Object Detection and Tracking Model

作　　者：蒋小惠[1] 曾小平[1] 

机构地区：[1]成都航空职业技术学院,四川成都610100

出　　处：《成都航空职业技术学院学报》2013年第1期45-48,共4页Journal of Chengdu Aeronautic Polytechnic

摘　　要：本文采用了一种运动目标存在情况下的背景重建算法,能较好地抑制外界环境变化带来的影响。该运动目标检测算法通过一定数量帧的学习、统计,找到各坐标点像素灰度等级的分割阈值。在对当前帧图像进行背景差分后,分割出运动目标区域。若存在运动目标,则根据串行边界跟踪算法获得运动目标轮廓。获得边界后,模型将绘制方框显示运动物体,并发出警报。该运动目标检测模型用Visual C++给予实现。This paper adopts the background reconstruction algorithm in case of moving objects. The influence of the change of external environment can be well suppressed ear be aceguired by the algorithm. Through certain amount of learning and statistics of frame, the segmentation threshold of the level of the pixel gray - scale of coordinates by the de- tection algorithm of moving objects. After the background subtraction of the current frame images, the region of moving ob- jects can be segmented, ff there is a moving object, the silhouette of the moving object can be got according to the tracking algorithm of serial boundary. Then, the model will draw the box to show the moving object and send the alarm. The model of the detection of a moving object can be implemented with the help of Visual C ＋ ＋ .

关 键 词：运动目标检测跟踪模型 背景差分 阈值 直方图 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]