📜 ⬆️ ⬇️

Video analytics: a pragmatic view

Does video analytics need a safe city? There is no doubt that the functions of recognition and search for events are necessary, but secondly, after the basic functions of the city video surveillance system are adequately implemented.
Today, urban infrastructure, including transport, utilities, and public institutions, is poorly integrated into the urban observation network and does not provide convenient access to high-quality video recordings to its main users — the FSB, the Ministry of Internal Affairs, the Emergencies Ministry, etc.


In my opinion, providing universal and convenient access to archival and live video is a higher priority task than the introduction of video analytics. On the one hand, universal access can be organized through the use of standard network video interfaces (ONVIF). The main advantage of the ONVIF interface when building a city video surveillance system is to standardize not only the streaming video broadcast protocol, but also the functions of search in the archive, transmission of alarm events and access rights.

On the other hand, universal access can be built on the basis of Internet protocols such as HLS and RTMP for transmitting video to browsers and mobile devices. This significantly expands the range of users of the urban video surveillance system.

For example, video of each entrance and playground of a multi-storey building may be available to residents of this entrance. At the same time, the massive use of some video analytics functions can now be justified in terms of technology readiness.
')
Moscow does not implement automatic monitoring of video signal quality using service detectors ( tampering alarm ). MGTS and other telecom operators incur significant losses due to fines for non-working cameras.

On the streets of the city can work reliably detectors of prohibited parking and crowds . You can configure the operation of these detectors on a schedule. Specialized detectors can be effectively used at transport facilities, for example, to detect falling of people on rails , movement against flow or running.

Nikolay Ptitsyn
Editor heading "Machine Vision"

Source: https://habr.com/ru/post/174309/


All Articles