printer friendly version

Intelligent video

Why is it an important development and how to best explore the opportunities presented by IV?

What is intelligent video?

Intelligent video (IV) has become a much talked about phrase for the network video and CCTV worlds over the last two years. Much of the talk has come from digital video recorder (DVR) vendors who have built intelligent video functionality into their devices to give them a new lease of life. But what is intelligent video and how can you tell the pretenders from the real intelligent video or 'IV' players? Intelligent video is about turning raw video data into actionable information and intelligence through analysis. Provision of this intelligence in surveillance (and other) video systems will support more informed, and therefore higher quality, user decision-making. Intelligent video is a natural extension of what Axis offers already in its network video devices. Building motion detection functionality into Axis' network video products several years ago meant that users of the company's cameras only had to collect and store images when a moving object enters an area of interest. Video management software can then be used to save 'security events', essentially periods of movement, for later analysis. The job of finding key video data is made that much easier.

Why is it an important development?

There are two key reasons why we see intelligent video as being a key development. The first is that it provides scope for network video infrastructure to be used for a whole host of new applications and enables existing security systems to be made even more accurate and effective. It is conceivable that surveillance systems deployed to trigger on motion could go further using IV functionality known variously as 'Tripwire' or 'D-fence' to demarcate a specific area within a field of view. Intelligent video systems can configure surveillance systems so they only collect video data when specific parameters exist and when anomalies to normal movements (say employees moving directly to their cars at the end of the working day between 4:30 and 5:30 pm) are detected.

The point is that systems can be set up to deliver far more targeted and specific information with IV. This means that in the future there will be far less reliance on manned guarding to spot developments which may have a security implication.

By configuring IV systems to specific parameters, it is also possible to get near instant alerts when pre-set parameters have been breached. These enable systems to help operators to work proactively. Timely alerts wherever they are (perhaps received on their mobile phone or PDA) will enable them to investigate further and ideally head off a full-scale security event.

What applications are end users likely to see first?

The first IV applications that are likely to find favour with end users are people counting applications. Traditionally, people counting is being used in the retail sector to count numbers of visitors to a store. These applications can be integrated with more sophisticated retail applications to study the success of a point of sale display or end of aisle sales offer in terms of 'dwell time' for example.

Network video systems can therefore continue to be deployed for multiple purposes: for business intelligence purposes helping retailers to increase sales and profitability through analysis of customer behaviour; for improving the customer experience through analysis of queuing times and observing reaction of waiting customers, helping support decisions to open new tills as queue times reach levels where they begin to detract from the customer experience. Network cameras can also be deployed to reduce shrinkage from till theft and spot incorrect change by triggering cameras to record on high frame rates during the periods that tills are open for example.

Another area where IV applications are likely to gain early traction are in road traffic monitoring and analysis.

What are the advantages of building IV capability into the network 'edge devices' themselves?

Some companies that are building systems for central monitoring stations and alarm receiving centres argue that intelligent video systems need to be built into management solutions at the point where the actual video analysis is going to take place. Axis takes a different view. It believes there is real merit from pushing intelligence to the edge device itself - making the analysis possible directly at the camera level.

Take the example of a licence plate recognition system again: it would make far more sense for a camera to be able to gather number plate images, turn them into data (a string of letters and numbers), cross check these with a driving licence authority's database and then transmit only the anomalies for further analysis, than simply transmitting all video data of cars going through a point on a road.

Costs of running that system would be considerably reduced as the bandwidth, network and storage resources required would be reduced significantly. In short, application developers will find it much easier to build a case for rapid return on investment through use of existing network and surveillance infrastructure by building IV functionality into edge devices including video servers and network cameras.

Conclusion

Intelligent video will undoubtedly fuel deployment of network video technology over CCTV and hybrid technologies. Leading industry analyst, IMS Research, recently revised its estimate for growth of the network camera market upwards to CAGR 53% (exceeding US$1 billion by 2009) partly, as they see it, because "network cameras will have embedded video content analysis capability to detect various behaviours or events". (Source: IMS Research News - Security & ID Newsletter March 2006). IMS Research, in line with Axis' thinking, predicts the first products with embedded IV capability will hit the market within the next 12 months. The IV market, in Axis' view, is set to experience rapid growth in 2008.

IV definitely offers a host of new reasons to use network video devices and also responds to end user desires to get more from existing network video infrastructure providing business intelligence which can be used to improve the effectiveness of organisations.

Share this article:

Further reading: