基于可靠背景模型的运动目标检测算法

Moving Object Detection Algorithm Based on Reliable Background Model

该文针对非参数背景模型的构建复杂和计算量大的问题,提出了一种简洁,快速和可靠的非参数背景模型构建方法和更新策略。首先通过MeanShift密度核函数来估计得到背景影响因素的估计值;然后通过对收敛点聚类,得到带权重的聚类中心;最后通过更新可靠背景模型和对运动目标轮廓提取,得到运动目标。实验结果表明算法可以在混乱运动对象的视频中提取出清晰的背景,能准确检测出运动目标,对光照。

In order to save the problem of the complex and computationally intensive non-parametric background model,in this paper proposes a simple,rapid and reliable non-parametric background model and updating strategies.Firstly,using MeanShift density kernel function to estimate the value of the background influencing factors;Secondly,through clustering the convergence points,and gets the cluster centers with its weights;Finally,by updating the reliable background model and extraction the contour of moving objects,gets moving objects.Experimental results show that the algorithm can extract a clear background of video without moving objects in the chaos environment,and can accurately detect moving objects with strong robustness of light,noise and a slight camera shake.