摘要
运动摄像机情况下的运动目标检测是视频监控中的难点和热点问题。为了能够有效地检测出运动目标,提出了一个基于SIFT(Scale Invariant Feature Transform)特征匹配和动态背景建模的背景差算法。首先利用SIFT算法提取特征点,采用RANSAC(Random Sample Consensus)方法求得仿射变换模型参数并实现图像的全局运动补偿,然后用背景差方法实现运动目标的检测,同时进行阴影和鬼影的去除。SIFT特征点匹配的准确性和RANSAC方法去除异常点的有效性使得仿射变换模型参数计算准确,动态更新背景模型的背景差则完整地检测出了前景目标。与Ninad Thakoor实验结果对比说明:该算法能够准确地检测出运动目标,并且保持了前景目标的完整性。
It is a difficult and hot topic in video surveillance to detect moving objects with moving camera. In order to detect moving objects effectively,we propose a background subtraction method based on SIFT features matching and dynamic background modelling. Firstly, feature points are extracted by SIFT algorithm to compute the parameters of affine transform model guided by RANSAC, and to realise global motion compensation. Then we adopt background subtraction approach to detect moving objects, with shadow and ghost removing. The precision of SIFT feature points matching and the validity of picking out outliers by RANSAC algorithm make the parameters of affine transform model to be computed accurately, and by the background subtraction approach with dynamic updating background model ,foreground objects can be detected perfectly. Experimental results demonstrate that comparing with Ninad Thakoor method, our algorithm can detect moving objects accurately and keep the integrity of foreground objects.
出处
《计算机应用与软件》
CSCD
2010年第2期267-270,共4页
Computer Applications and Software
关键词
运动摄像机
全局运动补偿
SIFT
RANSAC
背景差
目标检测
Moving camera Global motion compensation Scale invariant feature transform （SIFT） Random sample consensus （RANSAC） Background subtraction Objects detection
