24 September 2015

Could thermal imaging enhance security productivity? Enhancing the productivity of security personnel: the role of advanced imaging technology By Alex Schneider, Business Development Director (Imaging and Security), Acal BFi Security is a labour-intensive business. Ultimately, when a security threat needs to be investigated, deterred or prevented, human operators are normally required to attend the scene in person; for this most crucial element of a building’s or a space’s security system, human resources are absolutely necessary. In many other elements of a security system, however, machines can supplement, enhance or even replace the work of human operators, helping to increase their efficiency and productivity. High-performance video analytics, for instance, can cut the number of operators required to perform surveillance, and reduce the number of false alarms requiring a patrol call-out. Of course, the deployment of new technology costs money, and companies constantly face the dilemma of whether to bear the immediate cost of a capital investment for a return that will be enjoyed in the future. Now, however, thanks to recent advances in the technology of semiconductor components, an attractive security technology – the thermal imaging camera – has become markedly more affordable than it was in the past. Surely this is the opportunity that the security industry needs to dramatically increase its productivity, doing more with less, and doing it better as well? What is a productive security system? In fact, the application of the concept of productivity in the security industry is not as simple as it is in other industries. For a car manufacturer, productivity is easy to understand: if it improves factory operations so that its output rises to 1,100 cars a day, while employing the same number of workers which previously produced only 1,000 cars a day, it has improved productivity by 10%. But how can the ‘output’ of a security installation be increased? In fact, the output of a security installation such as a perimeter fencing system should be a negative: the total absence of intrusion events, because intruders have either been detected and their attempts thwarted, or because potential intruders were deterred by defences before ever embarking on an attempt. So the most ‘productive’ perimeter installation is one that achieves the objective – zero successful intrusions – with the smallest staff. Without a wholesale reinvention of the work of the security guard, the best way to increase staff productivity is to increase the effectiveness of the tools the workers use. And one of the most promising way to do this today is to replace conventional ‘visible’ (CCTV) cameras with the latest thermal imaging technology. Better prevention of intrusion A thermal camera increases a security system’s output – that is, it cuts the number of successful intrusion events – because it enables the earlier and more reliable detection of potential intruders. This is an inherent result of the differences between the nature of the infra-red (IR) radiation sensed by a thermal camera and the visible light radiation sensed by a CCTV camera. A thermal camera can detect IR emissions through leaves, fabric and other objects that obstruct visible light. So a potential intruder hiding from a CCTV camera behind bushes, for instance, will be seen by a thermal camera (see Figure 1). The thermal camera is also able to render meaningful images of intruders at a longer distance than a visible camera can. Providing fewer places to hide and earlier detection of intruders, the thermal imaging system will deter more intrusion attempts from being launched, and enable the earlier interception of those which do take place. A thermal camera also works perfectly in weather conditions that disable a visible camera: fog, smoke, haze, rain and snow are much less of an impediment to the operation of a thermal camera. So many intrusion events that would have been masked by these phenomena will be reliably detected by a thermal imaging camera. Doing more with less The thermal imaging system, then, produces a higher ‘output’ – and it can do so with fewer staff. That’s because thermal cameras complement today’s video analytics software far better than visible cameras do. Thermal cameras have very favourable characteristics: the image is stable and consistent at all times of day and night. By contrast, a visible camera’s output is strongly affected as the lighting changes from bright sunshine, to overcast conditions in daylight, and to artificial light at night. the outlines of potential intruders are clear, and unobscured by partial obstructions such as leaves on trees that break up the outline displayed by a visible camera the image is reliable. It is very difficult to fool a thermal camera, while camouflage techniques are used successfully by intruders to avoid detection by eye. This means that a video analytics system using images from a thermal camera will achieve far more reliable recognition of potential intruders’ body shapes. Crucially, it will then produce fewer false alarms. This means that operators can with confidence reduce or redeploy the personnel who are traditionally employed to continuously monitor CCTV display screens, and benefit from the reduction in the amount of time wasted by patrol staff on fruitless call-outs. Falling investment cost The technology of thermal imaging, then, might be highly effective. But is it not also esoteric and expensive, putting it out of the financial reach of all but the most lavishly resourced security operations? Not so. The semiconductor industry has performed the magic of increasing its products’ performance while reducing their price. So, where once a single thermal imaging camera would have cost more than £10,000/$15,000, units are now available at a cost of just £2,000/$3,000 (see Figure 2). These thermal cameras offer superior performance and longer range than visible cameras costing a typical £300/$450, so that security operators can often install one thermal camera where previously multiple visible cameras would have been required (plus high-power lighting for operation at night). As a result, the economics of surveillance have shifted, and the combination of the lower unit cost, superior performance and analytics compatibility have swung