抗干扰的运动目标检测算法  

Anti-interference Moving Object Detection Algorithm

作　　者：王明吉[1] 肖桂丹[1] 

机构地区：[1]大庆石油学院电子科学学院,大庆163318

出　　处：《计算机工程》2010年第3期232-233,236,共3页Computer Engineering

摘　　要：针对实时场景中运动目标检测的问题,提出一种改进算法。结合背景差分与帧间差分判定图像是否存在运动目标的过程中,引入形态腐蚀算子抑制大量孤立小噪声,采用基于矩形块的目标判定方法提高检测准确度,通过实时背景更新提高自适应性。实验结果表明改进算法可靠性高、抗干扰性强、背景适应性强。This paper proposes an improved algorithm aiming at moving object detecting in real-time scene. In the process of combing background differencing with interframe difference to dual decide moving object in the image being or not, it introduces morphological erosion operator to restrain a lot of isolated noise, uses the rectangular decision method to improve the precision of detection. Real-time update can also realize to enhance the adaptability of background. Experimental result shows that improved algorithm has high reliability, strong anti-interference and background adaptability.

关 键 词：运动目标检测 背景差分 帧间差分 背景更新 矩形块判定 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]