一种基于OpenCV实现的三帧差分运动目标检测算法研究  

Three-image Difference Algorithm for Moving Target Detection Based on OpenCV

作　　者：徐卫星[1] 王兰英[1] 李秀娟[1] 

机构地区：[1]河南工业大学电气工程学院,郑州450007

出　　处：《计算机与数字工程》2011年第11期141-144,共4页Computer & Digital Engineering

摘　　要：运动目标检测在视觉监控系统、医学图像分析、工业检测和军事等领域有着广泛的应用。运动目标检测处于这些应用的最前端,检测结果的准确精度对后续处理有很大的影响。文中提出了一种三帧差分运动目标检测算法,利用OpenCV函数库进行了仿真测试,并与基本帧间差分算法进行对比分析,结果表明该算法准确率高,运算速度快,能满足实时要求。Moving target detection is widely used in video surveillance,medical image processing,industrial detection and military field,etc.Detection of moving object is the first stage for these applications,so the accuracy of processing result influence greatly on the follow-up treatment.This paper presents a three-frame difference algorithm for moving target detection.Some simulation tests are carried out by the algorithm and the basic frame difference method,respectively,moreover the results are analyzed comparatively.The results illustrate that the proposed algorithm is more accurate and rapid,what can fulfill the realtime requirement.

关 键 词：三帧差分 运动目标检测 OPENCV 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]