城市安全监控视频图像目标自动跟踪仿真

Automatic Target Tracking Simulation of Video Surveillance Image in Urban Security Monitoring

对城市监控视频图像的目标跟踪,能够有效提高城市安全监控处理效率。对视频图像的目标进行跟踪,需要将采集到的视频监控每帧图像进行差分运算,计算出图像中包含的连通域的几何特征,完成视频图像的目标跟踪。传统方法定义摄像机视域,对时间与空间关系进行约束,但忽略了计算出图像连通域的几何特征,导致跟踪精度偏低。提出基于层次递归的城市安全监控视频图像目标自动跟踪方法。对视频监控图像进行预处理,将采集到的视频监控序列中的每帧图像进行差分运算,获得合理的灰度图像,给出监控图像背景缓慢亮度变化的鲁棒性,计算出视频监控二值图像中包含的连通域的几何特征和其数目,利用模糊信息推理机制给出图像目标自动跟踪最优策略,由此实现城市安全监控视频图像目标自动跟踪。实验结果证明,所提方法跟踪精度较高,为维护社会治安提供了可行的依据。

An automatic target tracking method for urban security monitoring video image based on hierarchical reeursion is proposed. After preproeessing monitoring image, we perform differential calculation on each frame of video in the collected video image monitoring sequence to obtain reasonable gray images. When the robustness of slow brightness variations in monitoring image background is given, we calculate geometric features and amount of connected domain contained in binary image of monitoring video. Finally, we use fuzzy information reasoning mechanism to give the optimal strategy of automatic target tracking of image. Thus, the automatic target tracking of monitoring video image of urban security is realized. Simulation results show that the proposed method has high tracking accuracy. It provides a feasible basis for maintaining social security.