printer friendly version

Intelligent technology is in your CCTV systems' future

CCTV systems have one fundamental flaw – despite putting up large numbers of cameras to maximise the coverage of an area, we seldom have anybody actually looking at much of the area for which we are responsible.

Sure, we can always go back and look at what happened, but many times this just is not good enough. It is a problem of logistics and economics. We cannot staff the control room with enough people to view the number of cameras we have available. So we have many of these cameras' views going to waste when it may be useful to view them. To make it worse, much of what is displayed on monitors will be uneventful, or sometimes even have content that is irrelevant to the risk profile of the site. This problem of coverage is not likely to go away soon. However, there are technologies being implemented now or in the near future that will change the risk coverage by CCTV significantly.

Intelligent technology that is being developed can help in a number of ways. Central to this is technology that can work with the operator in the background performing intelligent analysis of the camera views. Firstly, the computer can do things that computers are good at, and free the operator up to concentrate on other duties. For example, where automated recognition such as face recognition or number plate recognition can do the job, it frees the operator up to concentrate on the more relevant things that only people can do. In this way it is optimising the use of the system. Secondly, background sensing and alerts or alarms allow the detection of suspect conditions on cameras that are not currently being viewed by the operator. The computer can highlight these or bring these to the operators' attention for viewing and investigation. This could include certain types of movement or the presence of somebody or something in an area for defined periods. Through this, the coverage of the cameras has been made more efficient. Another option is acting as an early warning system. In this context, the technology can highlight conditions that could potentially be suspect such as objects left in an area, cars parked where they should not be, or excessive densities of people in an area. This technology is starting to become more common with advanced DVR systems.

Other technologies offer substantial potential, but have some development time still to go before they reach effective use. Where an operator is viewing a scene and needs to divert his or her attention to something else, parallel monitoring or tracking through tagging a particular object or person could allow operators to come back to that scene later and pick up more easily on the location and movement of the target. Building up an information base on targets, situations, and behaviours to use for risk management and future investigation is one of the most neglected aspects of CCTV. Use of an intelligent system that recognises behaviour and conditions can greatly help by increasing the amount and quality of information coming in. Technology can also help by simulating or verifying operator performance, either in contrasting the number of issues picked up by the intelligent systems, or the insertion of scenarios that should be picked up by operators and reviewing them against actual detection scores. Finally, intelligent technology can be used to extend the normal CCTV functions by building in algorithms that allow the detection of other things such as fire, smoke, production stoppages, overflows and so on that would add value to the CCTV function. I think the potential for intelligent analysis, sometimes called visual analytics technology, goes far beyond security-based CCTV to many forms of risk management and monitoring. The what, where, when, who, why and how will increasingly become available in our data retrieval system.

Besides the development needs that are required to realise some of the ideas for technology, there are some implementation problems with those we already have available. Cry wolf is when repeated warnings occur without foundation. Eventually, the people start disregarding the warnings and the potential of the technology becomes ignored. Use of blank screen-based alarm systems are a classic example, after 90 false alarms popping up on screen in an hour or two, the screens often get switched off because they are just seen as a distraction. We can also overload the operator with warnings on what could be happening so that they do not have the time to deal with everything they are faced with. Priority setting in such instances becomes important. We do not want CCTV to be reactive to only technology-based detection - people are still more efficient at being proactive and picking up dynamics and behaviour. If we prevent them searching for incidents, we will probably miss some of the most important. The timing of the event occurrence, or when we warn the operator is also something that needs to be considered. Too soon may just provide a lot of false alarms, too late may be exactly that - too late. Finally, where do we place the warnings within the display environment - does it become the centre of attention, or one of the tools the operator uses.

Technology has the potential to deliver higher performance in a number of ways:

* The level of work being performed - people can concentrate on the important and higher priority issues that need human interpretation.

* The volume of coverage of the system - all of these cameras are really working - not just providing a mass of information that will largely be lost and discarded. Clients are getting real capital return on their system investment.

* The quality of results delivered is higher, more things are being detected, and the base is being laid for a more intelligence and directed risk management process.

In five years' time, I think we will find intelligent systems on just about every worthwhile CCTV site. However, despite all the limitations with CCTV systems, the operator is still the key and final decision maker. The right operators make a significant difference to the system. Operators will continue to be able to do things that technology is unable to do and will do so for some years to come.

Dr Craig Donald, Leaderware

For more information, contact Craig Donald, Leaderware, 011 787 7811, [email protected]

Share this article:

Further reading: