序列图像运动目标的检测与提取  

COMPUTER & INFORMATION ENGINEERING COLLEGE OF INNER MONGOLIA NORMAL UNIVERSITY

作　　者：邹策千[1] 朱丽波[1] 王亮[1] 

机构地区：[1]内蒙古师范大学计算机与信息工程学院,呼和浩特010020

出　　处：《内蒙古农业大学学报（自然科学版）》2010年第2期281-285,共5页Journal of Inner Mongolia Agricultural University(Natural Science Edition)

摘　　要：本文对视频图像中运动目标的检测与提取的算法进行了研究,并在计算机上做了仿真。本文采用背景差分法,针对序列图像首先建立了基于高斯统计模型的背景模型,然后用差分法提取运动目标,用形态学滤波去除噪声。仿真结果表明,算法是有效的,能够得到较好的结果。A moving object detection and extraction algorithm for serial video images is studied in this paper,and also is simulated on computer.Using the background difference method obtain the object.First,established a background model based on Gauss statistical through several serial images,then differ current image with rebuilt background to extract the moving object,Morphology filter is used to wipe off the yawp.The simulation results indicated that the algorithm is efficient,and receives the better results.

关 键 词：图像目标 运动目标检测 目标识别 形态滤波 

分 类 号：TN911.73[电子电信—通信与信息系统]