An industry standard for CCTV video data interoperability

Abstract

The emergence of low cost, high performance, highly flexible and net workable CCTV cameras and related video signal equipment has created an urgent need for standardisation, without which end users are being faced with the prospect of difficult choices between incompatible systems and considerable confusion when it comes to long term investment decisions.

A standard is needed, not simply for signal format compatibility, but also to allow components from different manufacturers to exchange audio data, metadata, video analytical data as well as component capability information.

Early in 2008, three leading global manufacturers commenced cooperating with a view to persuading all manufacturers to work together in developing a new standard for networked CCTV interoperability.  This group, called the Open

Network Video Interface Forum (ONVIF) has made considerable progress and some parts of the new standard are available for comments.  Another group of manufacturers, called the Physical Security Interoperability Alliance (PSIA) has also developed and released some specifications, claiming its standard to be superior.

A single industry standard should bring considerable benefits to end users in terms of greater choice of equipment, greater flexibility easier integration of systems and lower life cycle cost.  It is not clear yet which standard will prevail.

Background

Analogue technology has formed the backbone of CCTV surveillance for many years and, since the technology used in those systems owes its origins to broadcast television, the basic signal formats have remained essentially unchanged for 50 years.  Video signal compatibility therefore, with some exceptions, not been a serious issue for users wanting to mix and match between brands of equipment.

Over the last 15 years however, with the increasing popularity and decreasing cost of remotely controllable pan tilt and zoom functions, even analogue systems experience compatibility issues when it comes to mixing equipment brands where camera control was needed.

The emergence in the last decade of digital CCTV recording systems that initially replaced recording devices such as video cassette recorders, did raise additional compatibility issues, but only when it came to recording formats.

In the early stages of IP (internet protocol) networks, camera outputs were sometimes digitally encoded for transmission over large distances but this was the exception rather than the rule because of costs and the limited bandwidths that were available.

Now, with the ready availability of gigabit networks, cheaper encoding equipment and cameras with highly sophisticated embedded capability, it has become feasible and cost effective for CCTV signals to be routinely transmitted on IP data networks which in most organisations already have extensive penetration.  Many high end cameras now have IP encoders embedded at little extra cost, meaning that they can be plugged directly into such networks, subject to security and bandwidth availability requirements.

Fully digital IP data networked CCTV has thus become an everyday reality, giving users a huge range of choices of equipment and capability to at least being able to interconnect their CCTV components using general purpose network infrastructure.

Why is there a Need for a Standard?

Data networks using internet protocol are already subject to standards in terms of the physical hardware and signalling:- i.e the cables, plugs and interfaces and the format of data packets which can be sent, but these standards do nothing more than ensure that a packet of data sent from one location can be received at another location and the original data retrieved.  In this respect, the network is no more than a 'pipe' down which the data is sent.

CCTV devices connected to a network need to speak a common 'language'.  Whilst this is not usually a concern for organisations who purchase an integrated system from a single manufacturer, confidence in the ability of devices to communicate and function usefully as a system (i.e. to inter-operate) is crucial when components come from different suppliers/manufacturers.

In order for a collection of CCTV devices connected to a data network to become a functioning CCTV system, the devices need to send and receive data that the other CCTV components can understand and interpret.  The kinds of devices might include:

• Cameras;
• Video encoders and decoders;
• PTZ domes;
• Recording/playback devices;
• Monitoring and control workstations;
• Other display devices;
• Analytics devices and software;
• Special purpose devices such as wireless interfaces.

The kinds of information that these devices need to exchange might include:

• The format of video data to be used (the predominant ones these days are JPEG, MPEG-4, H264);
• The format of audio data to be used (such as G.711, G.726, AAC);
• Data about the images (such as date, time, location, events of interest);
• Video analytical data (information about the contents of the image(s);
• The way devices identify themselves and the way they can find (discover) other devices on the network;
• The way the devices exchange and interpret information about their capabilities and the services they offer;
• The way devices can be initialised and subsequently controlled (such as PTZ controls, recording or playback controls or streaming) and which devices do the controlling;
• Security information that determines how the system is protected from unauthorised access or interference;
  The way devices can be re-programmed or loaded with different capabilities.

Benefits of a Standard

With an explosion occurring in sales of IP cameras, unless there is some form of standard language for CCTV devices connected to a network, it will remain difficult for end users to make use of the wide range of capabilities offered by different manufacturers.

The widespread adoption of a video interface standard will offer several advantages, including:

• New users and system designers will have the ability to mix and match from a range of vendors, thus assembling a system that most closely meets their needs;
• Greater choice for end users when selecting replacement video components because there will be no need to limit the selection to the brand of the system they own (provided it is compatible with the standard);
• Longer effective system life, hence lower life cycle cost, because it will not be necessary to replace an entire system if a manufacturer discontinues production of a key component;
• The prospect of sharing video feeds between different jurisdictions and organisations, with minimal hardware compatibility issues;
• Simpler installation and commissioning of systems;
• Reduced need for special software patches or other development costs when designing special purpose systems;
• Simpler, more uniform operating procedures because of a common configuration language;
• Greater variety in the products on the market, because smaller manufacturers will be able to produce compatible products.

Risks

There is a risk that some major industry players may opt out of the new standard or try to influence the development of the standard in a way that they think will give them the best market position.  The current ONVIF vs PSIA debate (see below) probably reflects this reality.

Standards can also have the effect of limiting innovation if they involve too many constraints on system design.  This could reduce the rate of development of new products.  Nevertheless, provided the standards are well conceived and flexible, the benefits are likely to well outweigh these concerns and the standard will have a useful working life.

There is some risk that users with recently purchased equipment that is not compatible with the standard may the see the useful life of their investment reduced as manufacturers shift their product development focus onto new standard-compliant products.  However, manufacturers may not compromise their clients in this way.

Moves Towards Standardisation

There are currently at least two industry groups actively working on standardisation initiatives:

1. The Open Network Video Interface Forum (ONVIF)
   In May 2008, three manufacturers (Axis Communications, Bosch Security Systems and Sony Corporation) initiated a new open forum "…aimed at developing a standard for the interface of network video products."...."The forum will be open for all companies who would like to participate in the standardization work…."
   
   A Technical Overview of the proposed standard was released by ONVIF in September 2008 and the first versions of the core specification and some of its component parts have also since been released for comment.  A meeting of all the technical committees is to be held in Tokyo on 3-4 March 2009 (refer http://www.onvif.org/ for copies of current drafts) and this will be the beginning of a continuous development process.
2. Physical Security Interoperability Alliance (PSIA)
   This rival group was founded in February 2008, with different but overlapping membership.  Its leading members include Cisco, Pelco, GE and Panasonic and it is considered to represent the interests of those manufacturers with a smaller market share.
   
   PSIA has already released some specifications (refer http://www.psialliance.org/) and has argued that its approach is "functionally rich and less expensive to execute and deploy than other architectures".  Some industry experts claim that PSIA has moved more quickly than ONVIF and that this is because it's standard will be easier and less expensive to implement.
   
   The outcome of the competition for dominance in the video surveillance standard is not yet clear.  Some commentators have suggested that both standards could coexist (perhaps analogous to the BHS vs. Beta battle of old).

Practical Effect of the Standards Development Process

The ONVIF and PSIA standards are voluntary industry standards, hence, like other industry standards they will not bind manufacturers to only produce compatible products, however, if one of the major standards organisations (such as IEC or ANSI) decided to take the matter on, there would be a stronger mandate.  Whether and when this will occur is still not certain.