摘要
针对静态背景下的视频运动序列,在研究现有的检测算法——帧间差分法与背景差分法的基础上,进一步研究了运动目标检测中背景动态建模的方法——混合高斯建模法,在此基础上提出了基于混合高斯模型与三帧差分的运动目标检测改进算法。由于使用背景差分法检测运动目标时,运动物体和阴影都将被看作运动的目标,于是研究了基于归一化RGB色彩模型的阴影处理方法,对阴影区域进行检测与去除。然后使用计算机视觉类库OpenCV结合Visual C++6.0对上述算法进行实现,取得了很好的检测效果。
Against the static background of the video movement sequence, based on studying of existing detection algorithm Inter-frame difference method and Background difference method, and further study of the moving target detection in the context of dynamic modeling approach- Gaussian Mixture Model, proposed an improved model based on Gaussian mixture with three differential detection algorithm of moving target. Then used OpenCV and VC6.0 to realized the algorithm and the detected result is very good.
出处
《电子测试》
2009年第9期86-90,共5页
Electronic Test
