Intelligent video architectures

DVRs and centralized intelligence

One solution for traditional CCTV systems being centrally monitored, is to take surveillance videos direct from the analog cameras into the IV-enabled DVR. The DVR will perform the intelligent analysis (people counting or car license plate extraction for example) before taking the remaining data, digitizing, compressing and recording it, and distributing resulting alarms and video output to authorized operators.

This approach is fine for systems that offer adequate capacity for uncompressed live video output to be transmitted to a central point. It is also appropriate for systems where the numbers of cameras are constant, as each DVR can only carry a specific number of cameras, and each unit is very costly.

Network video systems and distributed intelligence

A better and scalable alternative, compared to analog-based systems, is to rely on video servers attached to analog cameras locally to make the analysis of the video stream, and digitize and compress before passing information via the network for monitoring and storage.

The advantages of this approach are several. The first is that digitization and compression are done locally and then existing network infrastructure combined with the Internet itself can be used for low cost transmission of intelligent data. It is possible, for example, for only video data triggered by motion detection and accompanied by an alert message to be sent on to a centralized monitoring station for further action and analysis, possibly through use of more sophisticated IV applications. The load on the infrastructure and people involved falls dramatically.

A "distributed" approach enables intelligence to reside in "edge devices," as shown above in the network camera itself, or in the video server. The information can, at an appropriate time, be channeled to either a centralized server or to specific clients who need to conduct specific actions.

From a basic configuration, it is possible to join specifically compiled intelligence in order to govern operators tasked to a particular job. As organizations look to enhance their video systems, different types of distributed IV applications may be combined. The network distributed intelligence approach is limitless in scalability and thereby future-proof in regards to future expanding building blocks.

axisFlexSubContent