Open or Closed?
ONVIF standard extinguishes open debate
- By Oliver Vellacott
- Sep 01, 2011
The debate about so-called “open” and “closed” IP video surveillance
solutions has been ongoing for some time. Each side suggests that its
approach is better for the end user, with a reasonable amount of disinformation
being added for good measure. To a large extent, this debate
is being extinguished with the release of the ONVIF standard.
However, even with this standard, defining exactly what is meant by an “open”
system is still up for debate.
The ONVIF Standard
The global security and surveillance market is at a landmark stage as, finally, a
“standard” for IP video is really gaining traction in the market. ONVIF can boast
nearly 800 conformant products as of spring 2011, and it has clearly left the other
competing standard, PSIA, far behind. This was most apparent at an interoperability
demonstration at ISC West in April 2011. The ONVIF event was buzzing,
with more than 200 visitors, while the PSIA event attracted only a small number
It is clear that ONVIF is on the way to being the de-facto standard for IP-based
surveillance products, and open standards in the IP security market are here to
stay. ONVIF also is being incorporated into the International Electro-technical
Commission (IEC) standard for video surveillance systems and, once ratified, will
mean that ONVIF is a true international standard.
ONVIF provides the common language that enables IP security products from
a diverse range of manufacturers to work together. Rather than defining just a media
data format like H.264, ONVIF encompasses system control; for example, it
defines how the video management system asks a camera to start and stop a video
stream or move to a particular position.
Although initially focused on video, including cameras, analytics and NVRs,
ONVIF is being extended to include access control. If this is also adopted by the
market, it will be a substantial achievement, as the access control market is even
more proprietary and fragmented than video was.
As the market moves toward the adoption of an open standard like ONVIF, the
end user can expect to see a number of benefits:
- Increased flexibility and greater freedom of choice.
- More sophisticated and differentiated VMS features.
- Confidence that their investment is secure as it is possible to replace or upgrade
parts of the system as required.
- Less integration costs -- end users can choose the most suitable combination of
IP-based physical security products, regardless of vendor.
Because of these benefits, many tenders and RFPs are specifying that solutions
must be ONVIF-conformant. However, at this stage in the game, it must come
with a health warning. Plug-and-play interoperability of ONVIF products will
take time to achieve, as the standard settles down, and in the meantime there will
be some “teething” issues.
Prior to the emergence of ONVIF, multi-vendor solutions were based on proprietary
protocols, with a new driver needing to be written for each manufacturer’s
cameras -- costly and time consuming. In contrast, end-to-end solutions from
a single vendor provide a tightly integrated system, with components designed
to work together. The benefits are smoother project roll outs, optimum performance
and reliability, lower support and maintenance costs and a single-point
End-to-end solutions can also incorporate special features such as advanced real-
time analytics in the camera and bandwidth-saving capability that are not supported
within the ONVIF standard. Consequently, end-to-end solutions can often
outperform multi-vendor ONVIF systems. End-to-end may have strong benefits; however, in the past this came with a
drawback that users were locked in to a
Some providers of end-to-end solutions
are now opening up their software
to third-party cameras through ONVIF.
They are probably more likely to
provide robust ONVIF solutions earlier
as they will need to test and certify a
range of cameras to demonstrate their
openness, unlike the management software
providers who use the proprietary
drivers. One suspects because VMSonly
vendors might also prefer end users
to continue using the proprietary
protocols, their sales pitch is to support
any manufacturer, and ONVIF enables
everyone to make that pitch. The main
“raison d’etre” of the independent
VMS providers just disappeared.
The end user can therefore get the
best of both worlds by choosing an
end-to-end solution that also is ONVIF-
conformant. End users will be able
to mix third-party ONVIF cameras
with cameras that have been designed
to work in the end-to-end system.
This offers the benefits of a complete
integrated solution with the ability to
choose cameras from other manufacturers
to suit specialized needs.
Open Architecture -- Much More Than Just ONVIF
Being open is about giving people freedom
to choose. True open systems give
users choice in every aspect -- meaning
not just cameras but also storage, access
control, analytics and workstation.
Open also means giving access to the
inner workings of the system through a
software development kit (SDK). This
allows third parties to integrate with
wider systems such as process control,
oil & gas monitoring, EPOS and manufacturing.
An example of this is with
the integration of ANPR/LPR systems
with IP video. Specialist license plate
recognition cameras can be interfaced
to the network through video encoders
and thereby benefit from the distributed
architecture and “virtual matrix”
switching inherent in the best IP video
systems. ANPR/LPR processing engines
can also analyze video streamed
across the network from any camera in
Integration between security systems
is becoming an increasingly important
part of efficient security management
and incident response. With operators
in large surveillance systems having to
monitor hundreds of cameras, the use
of alarms from other systems to highlight
incidents and events is invaluable.
The best open IP video systems allow
tight integration with third-party systems,
of which the most commonly
used is access control.
Until access control adopts open
standards, such as ONVIF, this involves
the IP video manufacturer writing
interfaces for each access control
system in the same way as the original
pre-ONVIF “open” VMS systems did
for IP cameras. Integration interfaces
can also be written for non-security
systems such as BMS and electronic
point of sale (EPOS). Integrating EPOS, access control and video surveillance
together becomes a powerful
solution for applications such as casinos
and retail, where cash handling is
a major issue.
Cashier transactions can be stored
automatically with the relevant camera
video, enabling powerful post-incident
IP Video Architecture
When comparing IP video solutions,
there are many factors to consider. The
issue of cameras has now largely been
solved through the ONVIF standard.
This leaves other components such as
VMS and video recording. However,
one area that needs very careful consideration
is the system architecture. The
architecture has a significant impact on
the flexibility, resilience and scalability
of the solution.
Centralized server-based systems
have a single point of failure as all video
has to be transmitted through the server.
In large systems, the amount of video
streamed through the central server can
place significant demands on the network.
This means increased bandwidth,
latency, maintenance and costs.
In contrast, in a distributed serverless
architecture, video can be sent
from a camera to a local workstation
without the need to go through a central
server. NVRs also can be located
at the network edge near clusters of
cameras to reduce bandwidth and
improve resilience. A failure on any
part of the network will not cause the
whole system to fail.
Compression and Recording
The ONVIF standard defines how
video streams are managed and controlled
but not how they are compressed.
The way video is transmitted
within the ONVIF framework is
defined by video standards such as
H.264 and MPEG-4. These ensure
that any VMS or NVR can decode
any video stream that meets these video
standards. However, the quality of
the compression, i.e. the trade-off between
video quality and bandwidth,
can vary significantly depending on
how well the video compression standard
has been implemented in the
camera encoder. The amount of video
for a given resolution/frame-rate that
an NVR can store is very dependent
on how good the compression is in
each vendor’s camera. For the same
quality picture, one camera can compress
video 10 times more than another,
even though they both output
Calculating the amount of storage
required is an important part of the
costing and design of an IP video system;
in large enterprise solutions the
storage cost can be a significant part
of the overall cost. Vendors of end-toend
solutions typically supply storage
calculator tools that considerably ease
this process. However, in an ONVIFconformant
system, with cameras from
multiple vendors, the storage capacity
calculations become potentially much
The latest version of the ONVIF
standard (2.0) now enables development
of ONVIF-conformant storage
devices such as DVRs and NVRs. For example, a DVR maker could produce
a product that enables the video recorded
by that DVR to be played back
in any VMS that supports the ONVIF
standard for storage devices.
There are already storage products
on the market that are labeled ONVIFconformant,
but only from the camera
to NVR side. They can record video
from a range of third-party ONVIF
cameras, but a proprietary interface is
still required to review and playback
the video. Fully ONVIF-conformant
storage devices have not yet emerged,
but when they do they will have a profound
effect on the market.
In many cases, the recording solution,
whether it is standalone NVR hardware
or a software application, is an intrinsic
part of the VMS. ONVIF will enable the
decoupling of the storage from the main
VMS, providing increased flexibility and
a choice of storage, plus potentially lower
costs for end users.
Storage array manufacturers will
be able to produce standalone storage
products that can be integrated into
ONVIF-conformant VMS solutions
without the need for additional recording
servers. To date, this has been
achieved by the storage array manufacturers
pre-loading the recording
application software from the VMS
provider onto their appliance, thus
requiring them to create customized
products and to certify each VMS provider,
which is both costly and time consuming.
Because the ONVIF standard does
not define how recordings should be
stored on disk, just the interface, there
is an opportunity for differentiation of
products, e.g. optimized storage and
data retrieval methods, reliability, and
so on, giving end users the freedom
to choose products that meet their requirements
However, in addition to the benefits,
there will be challenges. For example,
setting up redundancy and failover
between storage devices from different
vendors could be complex, and the
ONVIF standard does not currently
support this. Today, the ONVIF standard
also does not provide a full configuration
interface, so a proprietary
interface from the storage vendor is
still needed for configuration. Until
the standard matures, these limitations
could be a barrier to ONVIF storage
devices appearing any time soon.
With so many of the big names in
the IP video market committed to ONVIF,
it has unstoppable momentum. It
brings wider choice and peace of mind
for the end user. But it is not by itself
the definition of an “open” IP video
system; it is just one of the many facets
that make up a truly “open” solution.
ONVIF also strengthens the benefits
of end-to-end solutions by offering the
best of both worlds: it installs easily yet
still gives freedom of choice.
ONVIF will undoubtedly bring significant
change to the market, which will
in the long term benefit the end user. In
the short term, there
will be a period of
“bedding down” as the
This article originally appeared in the September 2011 issue of Security Today.
Oliver Vellacott founded IndigoVision in 1994. He was previously a product manager with a background in intelligent camera products. He has a Ph.D. in electrical engineering from Edinburgh University.