IP video is more than just a change of communications technology: it opens up a vast range of possibilities.
The foremost driver of innovation comes from the removal of the first C in CCTV (closed circuit television). It is a common mistake to continue to refer to IP video systems as CCTV, when in fact the real excitement flows from the openness of IP systems. Once video (and sound and alarm inputs) is available as digital data, able to be routed and managed over LAN, WAN and Internet communications, it becomes many times more flexible with innumerable fresh applications.
The ever-increasing computational power that can be put into video components is another driver. These days far more video is being recorded than anyone could ever monitor or search, hence the next big trend is intelligent video. The advanced network camera can have built-in motion detection and event handling so the camera decides when to send video, at what frame rate and resolution, and when to alert a specific operator for monitoring and/or response. More intelligent algorithms - number plate recognition, people counting, face recognition etc - are being integrated into network cameras. Bandwidth is still, and probably always will be, a limited resource as demands for higher resolution and more complex data push requirements up, so being able to pre-process image information and transmit only data that matters will become more important. Intelligence at the camera level implies a far more effective means of surveillance than is possible with a DVR (digital video recorder) or other centralised system.
Megapixel imaging
The emergence of digital image formats made it possible to use image sensors with higher resolutions than NTSC/PAL (approximately 0,4 megapixels) to their full potential. The CCTV market has traditionally relied upon TV technology, but with flexible networking and digital image formats, this is no longer necessary. Security applications can now dynamically select the optimal format for the task. Unlike analog CCTV cameras and DVR systems, network cameras can deliver digital video in any resolution, format and frame rate, and can thus play a pivotal role in the new wave of high resolution, more advanced network video solutions and applications. They can, for instance, increase or decrease image resolution as needed.
High-resolution sensors will open up new possibilities on several fronts in the megapixel future. In addition to the familiar advantage of depicting more detail with greater pixel density, a less obvious development stems from the inherent flexibility of digital image formats. Not only will new sensors support 16:9 and similar formats; they will also be used to digitally pan, tilt and zoom, and to create multiwindow video. As resolution steadily increases and sensors become more sensitive, the use of traditional analog CCTV cameras and DVRs is decreasing in favour of network cameras and NVRs (network video recorders).
Complete adoption of network video systems will more than ever focus attention squarely on where video content is created: the network camera.
Video for all
As people continue to expand the scope of network video, we can expect monitoring and image distribution to expand to the general public and the home user. Webcams have been around for some years, but required a degree of specialised expertise to set up. IP video makes it possible to add cameras to systems and distribute video anywhere, almost trivially. Add to this the increasing availability of devices that can tap into the Internet, such as PDAs and mobile phones, and you have a creative cauldron.
On your way to a meeting and wondering what the traffic is like on the alternate routes? Accessing traffic-monitoring cameras from your desktop before you leave could let you see the traffic conditions for yourself. About to go from the restaurant to a club? Your mobile phone could show you the queues outside your favourite venues. You receive an SMS indicating an alarm in your home: you can immediately connect to cameras to see what is happening. All of these and more are already being done. What is changing is the simplicity and ease of installation as the technology becomes cheaper and smarter.
Intelligent video
With more video surveillance data being generated than ever before, how will we be able to use it all effectively? How will all the Petabytes (thousands of Terabytes) of stored data be classified and filtered, and how will video monitoring keep up? Video analytics, also known as video intelligence, will play a leading role in answering these questions.
The primary purpose of video intelligence is to provide better information, faster. It involves analysing a scene according to a set of rules to help identify and track actions, objects, or events. Such rules can cause a camera to notice a person falling, detect a suspicious piece of luggage, measure traffic, monitor a perimeter for intruders or even identify abnormal behaviour.
Video intelligence can be applied on different levels in a system and in real or recorded time. Traditionally, video intelligence systems have been deployed centrally in servers analysing incoming or previously recorded video data. This approach, however, does not adapt well to larger systems since the capacity of an industrial strength server is seldom more than eight video channels. As analytic algorithms become more sophisticated - demanding more CPU cycles - and video resolution and frame rates increase, system demands will outpace current capacity. This limits the cost-effectiveness and scalability of today's centralised video intelligence systems.
The trend is clear: network cameras can now handle most image analysis tasks themselves. And with powerful video intelligence invested in each network camera, system scalability is less of a problem and processing capacity is substantially cheaper. Such cameras also have the ability to directly analyse accurate digital sensor data at the source, before it is distorted by line interference and system delays.
The intelligent network camera concept also reduces bandwidth requirements, since video data is not transmitted unless needed. Most of the time, the analytic application will instruct the camera to only send meta-data texts, such as counters, licence plate numbers, and so forth. With these developments, the demand for open, standard-based systems will grow as users pursue complete, integrated, and efficient network video solutions. In addition to more advanced analytics, the new and flexible digital formats make it possible to put more video processing capability into network cameras.
In most scenes, what is usually of interest is motion or other changes. Unlike traditional CCTV, intelligent network cameras can respond by increasing the frame rate, shifting recording to local mode, increasing resolution etc. This series has mentioned numerous examples.
With any video surveillance system there is a concern about privacy. Video intelligence and network cameras can be put to work to alleviate some of these concerns. Unlike analog CCTV cameras that only send out one single video stream that anyone can look at, an intelligent network camera can encrypt and secure the access to the camera, as well as deliver several parallel video streams with different content and formats. Parts of the video can be hidden, or access to data limited according on the user's level of authorisation.
Video intelligence is expected to advance in waves of innovation in the coming years, giving rise to more powerful security and business applications. We will also see the emergence of a new product category as this happens: the truly intelligent network camera.
© Technews Publishing (Pty) Ltd. | All Rights Reserved.
printer friendly version