Video analytics server🔗
Video analytics server is a server exclusively dedicated to launching video analysis modules that process video data sent from other servers. Having said that, the video analytics server does not perform archiving of the video data being analyzed nor its broadcasting to client workplaces; these actions are performed on the general servers of a video surveillance system. Thus, the Video analytics server allows to remove the load associated with video analysis from the rest of the servers.
Pictures
Details
The Video analytics server supports all the video analysis modules apart from the following:
Loud sound detection
Fisheye dewarping
Warning
If the Video analytics server is operating under Linux, it will also be affected by the corresponding restrictions on the video analytics modules.
To assign a server as a Video analytics server, it is required to install a special license on it. At that, other licensing capabilities will become unavailable on this server.
Consequently, if the server was used as a Video Surveillance server before, it is required to transfer all of its cameras to another Video Surveillance server before using it as a Video analytics server.
Assignation of the server as a Video analytics server for a camera is performing in the camera connection settings of the Eocortex Configurator application.
Features and limitations
Warning
The Video analytics server must be of the same version as the video surveillance servers whose video it processes.
A single video surveillance server may have connections with several video analytics servers, and a single video analytics server can process video streams from several video surveillance servers.
If a video analytics server has been assigned for a camera, a motion detection will also be performed on the video analytics server; all the features that use the results of the operation of the motion detector, however, will continue running on the main server.
If a video analytics server has been assigned for a camera, all the video analytics enabled on the camera, with the exception of the unsupported modules, will be processed on the video analytics server.
The video analytics server does not perform recording of the archive, events or any other data related to video processing, apart from the logs that monitor operability. All the events and the archive are recorded on the main server.
To ensure proper operation of the license plate recognition modules and the modules that use neural networks, the corresponding components are required to be installed on the video analytics server. However, it is not necessary to install such components on the video surveillance server.
For the license plate recognition module to work, it is required to install the corresponding license protection key on the video analytics server.
The video analytics server cannot be set as main, backup, or replication server for the cameras.
The video analytics server cannot operate under Windows 7 operating system.
The main server cannot use a video analytics server if it runs under Windows 7.
It is not recommended to use the video analytics server as a main server in a multiserver system since it may lead to malfunctions of such system due to the fact that the video analytics server is under high load of the video analysis modules.
The operation of the archive on the main server when recording by the signal of a motion detector running on the video analytics server is described below.
When recording to the archive by the signal of a motion detector, the results of data processing on the video analytics server will be used.
In case the connection with the video analytics server is lost, the continuous recording to the archive will be enabled. The recording to the archive will stop as soon as the connection with the video analytics server is restored.
The continuous recording to the archive will also start if the video analytics server has not sent any events of movement or lack thereof in the course of 30 seconds.
Warning
If the connection with the video analytics server is lost, the video from the camera will not be analyzed. The software motion detection will not work either.