社区异常行为智能识别防盗报警系统仿真

当前社区智能识别防盗报警系统采用广角视频人工识别异常行为,不能实时对社区异常行为进行监控和报警,存在报警误报率较高、报警延时较长、准确率较低等问题。针对上述问题,提出基于社区异常行为智能识别防盗报警系统,介绍了智能识别防盗报警系统的总体架构,包括图像采集模块、异常行为检测模块、远程监控模块、防盗报警模块四部分组成,并结合运动目标颜色特征的粒子滤波算法实现了对社区异常行为实时监控和实时报警。实验结果表明,所设计的智能识别防盗报警系统,报警误报率较低、报警延时较短、准确率较高。

Community Abnormal Behavior Intelligent Identification Anti-Theft Alarm System Simulation

At present, the community intelligent identification burglar alarm system uses wide-angle image to artificially identify the abnormal behavior. But this system cannot monitor and alarm the community abnormal behavior in real time, resulting in high alarm false alarm rate, long alarm delay and low accuracy. Therefore, this paper proposed an intelligent identification burglar alarm system based on community anomaly behavior. Firstly, this research introduced the overall architecture of the intelligent identification burglar alarm system, including the image acquisition module, the abnormal behavior detection module, the remote monitoring module and the burglar alarm module. The particle filter algorithm based on motion target color feature was used to achieve the real-time monitoring and real-time alarm for community abnormal behavior. Simulation results show that the designed system has lower alarm false alarm rate, shorter alarm delay and higher accuracy.