基于音视频分析的区域安防管控平台

针对广场等区域人流量大、安保要求高的特点,设计并实现了一套基于音视频识别的区域安防管控平台.该平台从区域安防的需求出发,设计了嫌疑人员报警、人群拥挤报警和异常声音报警三大功能,以人脸识别、人群密度估计、异常声音识别三个音视频识别算法作为支撑,实现对区域内与人相关的异常事件的预警.首先介绍了基于深度神经网络的人脸识别算法,然后介绍了一种引入注意力机制的卷积神经网络模型来实现的人群密度估计算法,和一种基于多卷积神经网络模型融合的异常声音识别算法;最后介绍了平台需求与设计过程,主要包括平台建设的需求分析和界面设计.

基于雷达视频协同技术的铁路区域安防系统

基于对铁路周界管理要求及一线运行管理情况的深入调研,分析铁路安防系统现状和业务需求,提出基于雷达视频协同技术的铁路区域安防系统设计思路,搭建系统总体架构并实现系统目标功能。经过基础测试、入侵测试、现场试运行测试及环境适应性测试,确定系统的可靠性、功能符合性和实用性。测试结果表明,该系统在报警准确率和响应效率方面表现良好,可为相关管理人员提供便捷式应用软件及移动端管理系统。

Challenges of Tomorrowm——Fire Hazards of Systems Using Renewable Sources of Energy

Dramatic climate change, caused by over consumption of coal, oil and other traditional energy sources, as well as exhaustion of their reserves, imposed technological need to look for their substitution with new, renewable energy sources. The exploitation of these new forms of energy, solar, wind, earth and bio-fuels, initiated the development and application of new technologies, so far unused in practice. Rapid development and wide application of installations for use of renewable energy in many households and companies opened a whole new risk and danger in the fire protection field. With the purpose of introducing this problem to engineers in the area of fire protection, health and safety at work, this paper systematically presents various types of facilities for exploitation of renewable energy sources as well as potential dangers, risks and issues related to their safe operation.

城市轨道交通安防集成平台建设展望

随着《城市轨道交通公共安全防范系统工程技术规范》(GB 51151—2016)的发布,各城市地铁都在着手建设安防集成平台系统,但在实施过程中存在运营与地铁公安的职责和界面不清晰,与综合监控的功能重叠等问题。通过福州地铁5号线安防集成平台系统实施经验,梳理出各使用单位与系统功能的对应关系,并结合地铁公安的需求和线网规划需要提出相关建议,以期为后续安防集成平台系统建设提供参考。