The IP camera challenge

IP Video technology is now firmly established as a mature technology that brings significant benefits to CCTV surveillance applications.

To complement this, the growth in dedicated IP cameras has also been significant. However, unlike traditional analogue cameras where performance can easily be compared, IP cameras present a different proposition.

What is an IP camera?

An IP fixed camera or dome is a single integrated unit that contains the camera itself, codec for video compression and transmitter/receiver for the network.

All that is needed to connect the camera to the network is a single CAT-5 cable and local power, or just the cable if Power-over-Ethernet (PoE) is implemented.

IP cameras enable simple and easy deployment of networked CCTV systems.

IP cameras have been around for a long time, but generally they have a poor reputation for video quality compared to professional analogue cameras.

They originally emerged from the 'webcam' end of the market, aimed at applications like 'tourist spots' where a picture every 15 seconds is acceptable, so the commercial drive was for a low cost, low performance product.

One of the main limitations on performance is the compression technology implemented in these products.

Compression Technology

There are two predominant compression technologies used in IP cameras, MJPEG and MPEG-4 (see panel on page 45).

However, 99 per cent of IP cameras currently available in the marketplace are based on MJPEG.

MPEG-4 can deliver much higher video quality, but a proper implementation of MPEG-4 has until recently been too expensive to incorporate into a dedicated IP camera.

However, the cost, performance and power consumption of the camera sensor and MPEG-4 compression hardware has now reached a point where it can be packaged in a dedicated professional IP dome.

Domes are the obvious starting point for high-end IP cameras as the physical size of the enclosure makes it easier to accommodate the necessary electronics (see "Anatomy of a High End IP Dome", page 44).

The price/performance of the hardware has been further enhanced due to the investment by some IP video manufacturers in dedicated custom chips for MPEG-4 compression.

Video quality from these new domes when transmitted via a network and displayed on an analogue monitor is indistinguishable from analogue video directly connected to that monitor.

That's a long way from the early 'stuttering' performance of MJPEG IP cameras.

There can, however, be major differences in the implementation of MPEG-4.

It is possible to use MPEG-4 in 'I-frame only' mode, which is basically the same as MJPEG, and yet it is compliant with the MPEG-4 standard. In this case the IP camera can be labelled MPEG-4, but with video quality similar to MJPEG!

The problem with MJPEG is that it requires a huge amount of bandwidth to generate good quality video, typically 10-30 times what a good MPEG-4 implementation requires. This has an impact on bandwidth and storage.

The network provision has to be at least 10 times larger than it should be, and the cost of storage (often a major component of system cost) is also at least 10 times what it should be. Even though MJPEG based IP cameras are typically lower cost, the rest of the system is very expensive, resulting in an overall system cost that is higher than if a good quality compression technology is used.

Analogue cameras in IP systems

The professional end of the CCTV market has to date preferred to use standard analogue cameras connected to separate transmitter/receiver units to compress and transmit video over the IP network.

These units have far more processing power for compression and networking tasks, and therefore enable a cost effective deployment of storage and network bandwidth.

There are several other advantages to using this approach.

It is an ideal solution for the retro-fit market where existing analogue CCTV cameras are linked to new networks and IP video systems. This maximises existing investment and smoothes migration to a networked CCTV solution.

The range of analogue cameras available is also far wider than with IP cameras.

There is a camera for almost every need, and the advance in sensor performance is still continuing apace.

These advances generally don't feed through into IP cameras for some time. In fact the average price of analogue cameras has actually gone up in the last few years, as demand for high-quality video images has increased.

The CCTV system can only ever be as good as the performance of the front end camera sensor – "garbage in, garbage out".

That being said, IP cameras will clearly eventually take over from analogue cameras as they will drive down overall system costs through lower camera costs and reduced installation/wiring – one CAT5 network cable can replace local coax, 2-way audio, PTZ control and power cabling.

However IP cameras will not be accepted by the mainstream security market until the video quality is good enough, a diverse range of cameras is available and, most importantly, the system management software is available to support them.

It's a software game

With an IP camera the quality of the digital video displayed in the control room is a function of the camera sensor, compression technology and the software that is managing the IP video system.

Without capable enterprise level video management software the quality and benefits of IP video will not be realised.

These software applications provide control, administration and viewing of the live camera feeds and recorded video over the IP network.

The power of IP CCTV is not in the camera, but in how the management software delivers a useable application to the end user.

So how this software interacts with the IP camera, the recording hardware and the network is vital to overall performance.

This is a sea change from the world of analogue CCTV cameras, monitors and DVRs, where you could typically buy any product and evaluate its capabilities and performance in isolation.

With IP systems the solution is not a single box.

It's a collection of components which get distributed across an IP network, with the software to manage them as an absolutely essential component.

So IP cameras on their own are not enough; the real value lies in the software, and that should be at the top of anyone's decision making process when selecting an IP video solution.

One example of how IP cameras can benefit from the integrated approach is in the ability to have advanced features built-in to the camera itself, such as ACF and Analytics.

Activity Controlled Framerate (ACF) is designed to reduce network traffic.

If no movement is detected in the camera scene then bandwidth is dramatically reduced.

This feature is most effective in places where low activity occurs, such as in corridors, on fire escapes, or in buildings which are unoccupied at night. Recording in these areas becomes virtually free!

Analytics is the processing of video images to detect such events as congestion, stolen objects, cars parked too long outside a building, people moving the wrong way through security checkpoints, etc.

In integrated IP systems analytics can be run in two modes: real time within the IP camera, and post-processing, on any operator's PC via the management software.

The real time mode allows the system to automatically identify events as they occur.

Post processing allows operators to run many different scenarios on recorded video, such as looking for cars parked for more than two minutes, for example.

These two modes offer the best of both worlds, using analytics at the camera to identify events as they occur, and also providing advanced search tools for operators to analyse footage from a central location.

Human operators are particularly poor at watching video monitors for long periods of time, but are generally very good at confirming whether or not something is an incident, once it has been flagged up automatically by the system.

In summary, in an IP security system the value is in the management software – look first at that when evaluating a system.

That software, however, ultimately depends on the performance of the image sensor and quality of compression.

IP cameras still have a generally poor reputation for security applications, but the first few professional quality IP domes, that match or exceed analogue domes in quality, are starting to appear.