It will not have escaped your attention that we are now very much living in the High Definition – or HD – era. Look at all the HD satellite and cable channels, BluRay disks and HD TVs in peoples’ living rooms for the enjoyment of all that HD video.
Buying a new camcorder? Why wouldn’t you choose an HD model? Your mobile ‘phone probably has a two, three or maybe even a five megapixel camera embedded within it, while your ‘still’ camera will typically deliver eight, 10 or even 12 megapixel images.
The world of consumer electronics has truly embraced HD and the benefits of megapixels. As product volume has increased, so costs have come down in line. What, though, does all this have to do with CCTV?
Megapixel security cameras have been around for a few years now, supplied by specialised vendors such as Arecont, IQinVison and Luminera and via a few larger brands like Axis Communications. Nevertheless, megapixel cameras have so far accounted for just a few percent of the total camera market. In other words, megapixel CCTV has not enjoyed the same dramatic rate of market adoption as has been seen in the consumer world.
That said, things are changing. Megapixel CCTV has reached a tipping point. Simply put, all the key technologies needed for megapixel CCTV are advancing rapidly to the point at which the resulting surveillance system will be significantly more attractive than analogue or 4CIF IP camera systems in an increasing number of end user applications.
To understand why, we need to think about the key elements of the surveillance system and of the cameras themselves – lenses, image sensors and compression technologies – and the management systems that make it all accessible. More of this anon. First, why should end users care about these developments?
Benefits of megapixel imaging
The key characteristic of megapixel IP cameras is their ability to provide substantially higher resolution images than conventional cameras. A high quality 4CIF IP camera delivers images of 704 x 576 – around 400,000 pixels.
Today, standard megapixel products range from 1.2 to five megapixels, with some specialist formats exceeding 11 megapixels (extending up to 21 megapixels for more ‘exotic’ line-scan cameras). This extra resolution may be used in two main ways – to see more detail from the same field of view, or to obtain a larger field of view with the same level of detail.
To illustrate the value of this, consider a two megapixel camera with a typical resolution of 1600 x 1200. Assuming a lens is selected that provides an equivalent field of view, the megapixel camera will be able to deliver more than twice the detail in the image. If your priority is area of coverage and you don’t need more detail, then you could select a camera position and lens that lets you use a single megapixel camera instead of two or more conventional cameras.
Higher levels of detail are proving useful to help operators recognise people, read number plates on vehicles or identify logos or other marks on property.
In comparison with PTZ
There’s also a useful comparison to be made with PTZ cameras. With a PTZ, you zoom in to see more detail by remotely adjusting the lens. The image you see – and any recording – will only be of the close-up image. With a megapixel camera, the end user can zoom in digitally, choosing to view only a small part of the image. As you’re not adjusting the image that’s being captured by the camera, you can simultaneously look at the whole image and continue to record everything.
Why is that useful? Well, as a security manager recently remarked to me: “The limitation of a PTZ is that it’s always pointing in the wrong direction”. What this practitioner meant was that when you’re concentrating on a detail in one place, the camera cannot be looking anywhere else. With megapixel cameras, you can always see and replay the full image (or any detail contained within it).
Megapixel cameras can also provide installation and maintenance benefits. In applications where one megapixel camera replaces two or more conventional cameras, there’s the saving in poles or brackets and the commensurate saving in cost and time spent on installation.
Megapixel cameras may also be powered over the same network cable that carries away the image stream, avoiding the need for mains power having to be made available at each camera position. In digital PTZ applications, of course, there can be maintenance and reliability savings since megapixel cameras have no moving parts.
Technologies at tipping point
As stated, when examining megapixel imaging there are three main elements – the lens, the image sensor and video compression, all integral components of the camera – combining with a fourth (the networking and management software that delivers images to security managers and CCTV operators) to deliver a viable solution.
The first key technology needed for high quality megapixel images is the lens. It’s important to recognise that the megapixel count is not the only factor in determining image quality. Again, to compare with the consumer electronics market, a photographer buying a Top-of-the-Range 12 megapixel digital SLR camera will never achieve the image quality the camera is capable of if they opt for the standard ‘kit’ lens supplied in low cost bundles. In the hands of a professional, the same camera (sensor) equipped with a high-resolution ‘professional’ lens will deliver markedly superior images.
A megapixel camera requires a higher resolution lens than ordinary CCTV cameras, but these are still relatively expensive for normal CCTV lens formats (such as C/CS). However, as the volume of megapixel lenses sold begins to increase, we can expect market forces to substantially reduce costs through economies of scale.
Where analogue cameras commonly use Charge-Coupled Devices (or CCDs) to convert light into an electronic signal, IP cameras generally use small Complementary Metal Oxide Semiconductor (CMOS) image sensors – the same technology as in digital still cameras.
So far there have been two main limitations in this part of the system – limited low light performance and dynamic range. Until quite recently, CMOS sensors of the scale used in megapixel cameras have not offered the same low light performance as some conventional resolution cameras (as a direct consequence of delivering higher resolution). They’ve also exhibited a limited ability to accurately capture detail in both very bright and very dark areas of a single scene simultaneously.
In the conventional camera market, recent years have seen considerable successes in overcoming similar limitations, and we can speculate that similar improvements could soon be made as sensor and technology vendors who have solved the problem there apply their knowledge in the megapixel realm.
Is bandwidth an issue?
Some technology vendors will claim network bandwidth isn’t an issue as you can use a dedicated IP network to carry your megapixel CCTV images: the equivalent of the coax cable infrastructure for analogue cameras. To some extent that’s true, but the data rates and sizes of present generation cameras require careful planning and evaluation of the total cost of the solution – including storage – to make sure the end user is really deriving the benefits they’d expect at a cost they can afford.
The technology that substantially controls the amount of data involved is the video compression system in the camera. Video compression is used within the camera to convert the raw electronic image into data that, in turn, may be efficiently transmitted over the IP network and stored in an NVR.
Currently, one or two megapixel cameras are using MPEG-4, or proprietary compression systems (an extension of current compression systems). So far, the required cost, power and size envelope has really limited compression for two megapixels and upwards, so MJPEG is typically used to stream a series of compressed still JPEG images… but this is changing.
The security market is now benefiting from consumer demand for higher quality video images in mobile telephones and HD camcorders. This has been driving up the performance of semiconductors and forcing their costs down. The HD home video market has standardised on H.264 compression at around two megapixels. This market-led decision is now filtering through to security.
This concerted drive forward was evident at the ISC West and IFSEC Exhibitions earlier this year, with exciting announcements regarding the latest megapixel camera technologies from vendors such as Arecont.
The key change here is that the increased efficiency of the compression system provides a substantial reduction in the data rates required for megapixel images. In turn, that renders real-time frame rates a genuine possibility at useful resolutions of two megapixels and above. Now, semiconductor vendors TI and Micron are planning to move their reference designs to H.264.
Reducing the network’s burden
A change of this nature reduces the burden on an IP network, but also – and this is crucial – makes the cost of storing megapixel images viable in real-world projects requiring anything between 14 and 30 days of image retention. It’s a big change from MJPEG megapixel images where, at present, it’s common to compromise on frame rate and storage retention in order to obtain the benefits of higher resolution.
Of course, HD surveillance is not solely about the megapixel cameras themselves, but also the software and management infrastructure that’s in place to support them. To this end, the adoption of megapixel imaging technology has been made easier by the fact that IP infrastructure is now the universal standard for networking. As a result, there’s both pressure and indeed opportunities for security teams to use and benefit from this enterprise-wide investment.
The benefits of an IP-enabled, end-to-end intelligent digital video solution are well documented: real-time analysis, movement of video around the network, centralised recording and management of multiple sites to enable automatic threat detection, instant verification, event resolution and effective investigation among them.
Towards a second generation
Early adopters have added megapixel cameras to their surveillance operations to work alongside conventional PTZ, legacy analogue and regular IP cameras in delivering the best mix of imaging technology.
That technology is managed over an enterprise-class, scaleable and open IP system that encompasses edge devices (including third party edge devices), NVRs, video content analytics and Control Room visualisation.
During the next six to 12 months, major advances in the second generation of megapixel cameras will be arriving on the surveillance scene, delivering the technology innovations and cost efficiencies of megapixel imaging at a price point that will make them a compelling proposition for clients.