视频监控运动目标图像优化检测仿真

Image optimization and detection simulation of moving targets in video surveillance images

针对传统的ViBe检测算法在检测中出现鬼影、静止目标、阴影的缺陷,提出将像素值转化到小波域中进行分频,经过算法的优化设计,实现一种基于小波域中检测的新算法。针对ViBe算法中存在鬼影和静止目标的缺陷,将其背景更新策略设计成阶梯型更新,并赋予前景点也具备传播更新背景模型的特性,从而克服了ViBe算法中的鬼影和静止目标缺陷;接下来,利用小波分析的多尺度空间信息进行算法设计,从而达到去除阴影的目的;最后通过3组视频序列进行综合实验,用3个标准进行评价实验结果。实验结果表明,该算法实现了运动目标的有效检测,且优化了ViBe算法的缺陷。

In allusion to the defects of ghosts,stationary targets and shadows appearing during the detection of the tradi tional ViBe detection algorithm,it is proposed that the pixel values are converted into the wavelet domain for frequency divi sion,and a new algorithm based on the wavelet domain detection is implemented after optimization design of the algorithm. In al lusion to the defects of ghosts and stationary targets in the ViBe algorithm,the background updating strategy is designed into lad der type updating,and the foreground points are also given the characteristic of propagating and updating the background model, so as to overcome the defects of ghosts and stationary targets in the ViBe algorithm. The multi scale spatial information of wave let analysis is used to conduct algorithm design,so as to achieve the purpose of shadow elimination. Comprehensive experiments were conducted by using three sets of video sequences. The experimental results are evaluated by three criteria,w hich show that the algorithm can achieve an effective detection of moving targets and optimize the defects of the ViBe algorithm.