SBRI: innovation in rail security surveillance analytics, phase 1

SeeQuestor是一家在视频人工智能领域处于领先地位的英国中小企业，在提供创新和获奖产品方面有着良好的记录。SeeQuestor将实现其对“智能CCTV(ICCTV)”的愿景。先进的分析平台将获取录制的视频流，并通过SeeQuestor的元数据引擎对其进行处理，使数据可以立即进行搜索，同时向Network Rail的观众呈现感兴趣的事件(就像实时一样)。主要功能包括：威胁检测(如枪支、刀具、遗弃的袋子)；行为分析(游荡、奔跑、异常活动)；人群分析；即时定位感兴趣的人等。该示威者旨在促进快速事件反应，以及提前预测可能影响公共安全的事件。利用尽可能多的Network Rail摄像头基础设施，并设计为易于扩展，包括用于非常大的摄像头网络。ICCTV将提高网络铁路闭路电视操作员的技能，通过提供对数百个摄像头的实时视频进行快速、智能和同步监控，这是任何人都无法做到的。因此，通过将运营商的注意力集中在最有效的位置来极大地提高监控覆盖率，通过主动检查和识别感兴趣的人和行为以及接收相关的安全信息(对人员安全至关重要)来缩短响应时间，以及改善乘客服务和体验。然而，SeeQuestor今天能够演示其中的一些功能，该公司正在申请资金，以进一步开发、改进和生产其算法，并建立一个强大的软件演示平台。

SeeQuestor - a leading UK SME in the field of video artificial intelligence with a track record of delivering innovative & award winning products - will deliver its vision for "Intelligent CCTV (iCCTV)". The advanced analytics platform will take recorded video streams and process them through SeeQuestor's metadata engine, making that data instantly searchable, whilst presenting events of interest to Network Rail viewers (as if in real time). Key features will include: threat detection (e.g. guns, knives, abandoned bags); behavioural analysis (loitering, running, abnornal activity); crowd analysis; instantly locating persons of interest & more. The demonstrator is designed to facilitate rapid incident response as well as to predict events that might impact public safety ahead of time. Leveraging as much of the Network Rail camera infrastructure as possible and designed to be easily scalable, including for very large camera networks. iCCTV will enhance the skills of Network Rail CCTV operators, by providing rapid, intelligent and simultaneous monitoring of live feeds from hundreds of cameras, something outside the capabilities of any human. Thereby vastly improving monitoring coverage by focusing operator’s attention to where it is most effective, improving response times through proactively checking & identifying persons of interest and behaviour and the receiving of associated safeguarding information, key to the safety of personnel, as well as improve passenger service and experience. SeeQuestor is able to demonstrate some of these functionalities today, however, is applying for funding to further develop, refine & productise its algorithms and to build a robust software demonstrator platform.