Want the best VCA there is? Connect the CCTV system to our brain!

brain-wiredToday a human being is filled out with the best in the world availliable VCA system. Our eyes see pritty good during day and night times and our brains will take care of a mostly perfect content analysis! The system works so good that rain, snow, leaves, flags, spotlights (dark/light), animals, crowling people etc et etc, is perfectly detected or filtered out!

Let’s be honest to eachother, Isn’t it true that we do not mention that leaves are moving when it is windy? Isn’t it true that we perfectly see the difference between a dog or a human beiing who behaves like a dog? Our human brain is developped to recognize these kind of situations. It is prepared to take care of dangerous situations and helps us to react as fast as we can on it. We know how smoke looks like and we know that we have to take care of it.

Imagine that we could connect our today’s CCTV systems on a human brain. We then would have simply the best VCA system there is in relation to the today’s on the market availiable Day/Night CCTV systems. In Lausanne (Swiss) the Blue Brain project is located. This project is based on a human brain replica. Within 10 years, the project leader expects that the replica brain is ready and will function. Nowadays to represent the total human brain, we need a capacity of 12 football fields filled out with super computers and a huge amount of power…

Funny isn’t it that the best VCA system is as big as a football and only needs a sandwich with peanut butter 😉


Future or reality?

mazda-kaan-e1-future-carImagine this,

you are watching a tv show or a movie for example. In the movie a real nice car is presented to you. How nice would it be then if you virtually could click on it to get all the details? How nice would it be then if you automatically would be forwarded to the manufacturers website or even better,… Imagine you are watching the show on your mobile device. Let’s say you iPhone for example 😉 You see the car, you click on the screen at the car’s position and the software directly will show you the nearrest dealer based on your build in GPS coordinates,… open up your also installed navigation software and guide you there,….

vca-for-videoImagine that the show is about fashion. You just click on the dress or suite you want to be informed of and you will get an up to date overview presented with the prices and stores near you (based on the build in GPS).

Video Content Analysis is still under development. Companies like Intel are working together with the movie industry to develop the above mentioned Video Content Analysis techniques. Within a few years these techniques are no future anymore! Within a few years it is possible for us to explore our movie, Blue-Ray, TV show or any type of video based content just via a simple click.

It is not about what video brings you, it is about the experience it brings!

Kind regards,



standaardisatieWhile we all speak lovingly about the goodness of standards, the fights to set and control standards can be severe. Indeed, a battle may be emerging in the road to setting IP camera standards.

I see 3 main participants in this process (listed alphabetically):

  • ONVIF: Lead by Axis, Bosch and Sony, they seem motivated to protect the interests of the largest selling camera manufactures
  • PSIA: Lead by Cisco and supported by a half dozen camera manufacturers, they seem motivated to protect the interests of manufacturers with lower IP camera market share
  • SIA standards committee: the oldest of the 3, this committe has actually published standards and looks to be the least political (though not an industry alliance like PSIA and ONVIF, SIA could be the organization that eventually manages the process of standardizing the winning specificaiton)

Why is this Happening?

The IP camera business is exploding.  What may be a $700 Million USD market today is poised to be a multi-billion dollar business in a few years. Indeed, the fight for standards is really a fight to shape a market that will generate sales of $20 – $40 Billion USD in revenue over the next 10 years.  With the growth rate and such large revenue at stake, it is reasonable to see why companies would be motivated to act.

ONVIF and PSIA were launched publicly early this year.  This week, they both issued new releases (ONVIF’s release and PSIA’s release).  Both organizations will be making a push next week at ASIS.  While SIA has the infrastructure in place to create ANSI/OSI standards, large manufacturers are certainly motivated to try to shape and expedite development in their favor.

What is the impact of standards?

Customers win because standards ultimately drive down costs by eliminating market control over interoperability.  You can make any IP camera work with any IP video management system. The problem is that it’s often custom and costs $10,000 – $20,000 USD or more to do so. This makes it infeasible for most.  Once you have standards in place, this cost diminishes, allowing many more suppliers to be supported. With more suppliers in the market, incumbents usually have to respond to the pricing pressures of new competitors.

As such, standards can actually undermine the market leader’s position.  Market leaders generally do not need standards because most parties already support their products.  For instance, Axis is already supported by almost every IP video management system.  A standard would not increase their support much at all.  However, it could dramatically increase support for competitors (which removes a key advantage that Axis currently has).  On the other hand, standards tend to make the market grow quicker (but usually with downward pressure on pricing).  As such, market leaders may be able to benefit from standards but it is certainly risky.

On the contrary, companies with small market shares or those trying to get into the market benefit greatly with standards.  One of the biggest problems for analog camera companies trying to get into the IP camera market is that hardly no one currently supports their IP cameras (an ironic state of affair for analog market leaders like Pelco, GE and Honeywell).  With standards in place, this would eliminate a key barrier and allow them to compete with an Axis on other dimensions they might be stronger at (like distribution or imaging quality, etc).

Overviewing the 2 Key Alliances

So it seems to me that you have ONVIF vs PSIA (you can also consider it Axis vs Cisco or Leaders vs Laggers).

ONVIF: Since ONVIF represents Axis and Sony (with about 45% market share of IP cameras), they should be motivated to control or delay the adoption of standards.  Since they have the market power, they do not need to rush into any program that does not support their interests.

PSIA: They represent the companies with smaller IP camera share (Panasonic, Pelco, Honeywell, GE, DvTel, Verint, IQinVision – all together are a fraction of Axis and Sony’s IP share) and Cisco.  Cisco is an interesting factor here because Cisco is more motivated to sell networking equipment than cameras.  Cisco has a heavy incentive to make cameras as cheap and easy to use as possible because that sells more of Cisco’s core networking products (indeed, this ironically may make Cisco the best consumer advocate for IP camera standards).

Read an in-depth comparison of PSIA and ONVIF including observations on who is likely to win.

written down by John Honovich

The working area: Lane Assist, a proven technology based on iVCA

Before I wrote you about the level of expectation. Based on today’s experience a lot of customers aren’t satisfied about the today on the market presented iVCA solutions. But what if we focus on for example the automotive industry. Companies like Bosch or DENSO Products develop very advanced solutions which have an already proven state of succes. Isn’t it strange that a lot of us do not believe in iVCA solutions but that we do trust these solutions when they are filled out in our car? Why is it that we let steering our car by an iVCA integrated solution?

To give you an example of a proven technology in this part of market segment we will take you with us at a journey guided by the DENSO website:


The lane keeping assist system detects lane markers on the road, and assists the driver’s steering to help keep the vehicle between lane markers. When the system detects the vehicle straying from its lane, it alerts the driver visually as well as with a buzzer, while applying a slight counter-steering torque, trying to prevent the vehicle from moving out of its lane.


DENSO developed two key components for the system including:

  • Vision sensor, which detects shapes and positions of lane markers.
  • Steering assist electronic control unit (ECU), which calculates a target steering torque based on the data from the vision sensor, and then sends a steering torque signal to an electric power steering (EPS) ECU to control the EPS motor. The steering assist ECU further determines if it is necessary to alert the driver of the vehicle’s deviation from the lane with a buzzer and a display.

DENSO integrates the steering assist ECU with the vehicle distance ECU for the adaptive cruse control system and/or the pre-crash ECU for the pre-crash safety system.

For more information visit the DENSO website. Click here to visit them…

Belgian Made Video Analysis Revealed Marketable in Kuwait

A coast surveillance project successfully realized by ACIC

pic3Innovating spin-off of the research center Multitel, the Polytechnic Faculty of Mons and the Catholic university of Louvain, ACIC gets promoted after developing a panoramic surveillance software based on mobile cameras.

After a test-project successfully achieved in Kuwait (the surveillance of a highway stretch along a US army base), ACIC was asked to fulfill another challenge : watching a 10 kilometers coast running alongside to military base and oil refineries.

An alternative to the radar operating since June 2009

To offer a detection tool able to rival radar, working 24/7, the use of PAN-TILT (mobile) thermal camera was inevitable.  The individual sequences brought together in a panoramic picture are submitted to the program named MvPanoramaDetection.  As an intrusion is revealed  the software orders the camera to follow the intruder for a few seconds to allow its identification but no longer to avoid other potential intruders.  Therefore its scanning rhythm is especially based on sea transports velocity.  The precision of the system permits the identification of a zodiac 300 meters far from coast, or a boat far from several kilometers.  Successfully delivered on last June, the system is fully operational.  In conclusion, the MvPanoramaDetection represents a less expensive solution than radar, giving the security members the advantage of a visual identification.  Well done!

ACIC’s Kuwaiti project in figures…pic5

9 kilometers long and 3 kilometers wide area, 3 oil refineries and 1 military base, 3% of worldwide oil production, 800 cameras of which 50 are mobiles, 15 cameras equipped with MvPanoramaDetection, 160 cameras equipped with MvActivityDetection, 19 servers in 7 places, 24/7 surveillance, detection from 300 meters to several kilometers pic4

Few competitors for that technology looking for partners

This was not the first experience for ACIC in Kuwait.  UNO had already chosen ACIC’s solutions to watch over their new building opened recently in Mishref.  With MvPanoramaDetection, ACIC definitely reaches a higher level in its expertise, meeting their major concurrent among Israeli and American companies.  More locally they keep developing and promoting their other software’s (MvPeopleCounting, MvActivityDetection, MvCameraAnomaly…) in Belgium and abroad.  As their solutions are adjustable on most existing surveillance systems, ACIC is constantly looking for partners, integrators and distributors to sell or install its highly technologic solutions.  Seeing how satisfied clients are lets think ACIC is ready to open to new horizons for coming years…

About ACIC

Created in 2003, ACIC’s team is made of engineers, computer scientists and Ph.D.’s in applied sciences specialized in the creation of video analytics systems.  A camera equipped with an  ACIC solution become an intelligent and powerful detector able to reveal suspicious-looking  presence, abandoned parcels, thefts (something missing in the image) or someone making wrong way.  Following a pre-defined procedure, the system can generate alarms or transmit data’s to the user, like it’s the case with MvPeopleCounting.

ACIC’s advantage: a young and capable team, with great skills of adaptability and creativity helping them to create video analysis software for very specific destinations and use.

Through MvPanoramaDetection, the whole work of ACIC has recently been rewarded by Agoria’s Bravo L’Industrie Award.


Which will be the best?

winner-themeA lot of customers, but also manufacturers ask me “Within your expertise, which brand is the best?”

The first time this question was asked to me I thought, “Cool… I never looked at it this way”

In my opinioun there is no best brand or let’s say a winner. As discussed in my other article, we need to deliver solutions which fit the markets (customers) needs. Of course a specific brand could better fit these needs then an other brand but still that doesn’t mean that it is the best brand.

Video Content Analysis based systems are mostly developped within a specific market segment. Face recognition, traffic control or for example smoke detection are very different types of situations and because of that, these systems needs to be filled out with specially for these situations developped solutions.

You can imagine that for me it is not possible to answer the question. It is a pitty to see that most of the people are dissapointed by that. In their opinion, I as an expert have to answer it. If I can’t answer it,… what does it mean then?

The fact that answering this question isn’t possible at this moment just shows us how broad the term VCA or IVA is used.  Splitting up the market into various segments would make the question more simple to answer but also then I am convinced by the fact that there are no really bad brands but that it just depends on the combination of the right brand used in the right situation.

As I mentioned before,… we have to take care of managing the customers level of expectation.

Kind regards,


Observision receives exclusive certification for its video analytics software

Xtract™ system in UK Government’s catalogue of approved security equipment
tkh_buttonHeteren, the Netherlands – July 20, 2009 – 
Observision BV, provider of the fastest Video Content Analysis system for Video Surveillance,
today officially announced that it has been granted an i-LIDS certification for its video analytics software
“Observision Xtract™ system 3.2” by the independent UK Home Office Scientific Development Branch (HOSDB)
as primary detection system for operational alerts as well as event based recording in sterile zone monitoring
applications. The certification implies a mention in the honourable UK Government’s catalogue of approved
electronic security equipment and is usable in- and outside the United Kingdom. 
Observision’s Xtract™ system 3.2 now belongs to a select group of video analytics technologies that has been approved by the
HOSDB for governmental use. Mark Bloemendaal, CEO and co-founder of Observision BV: "We are proud that we have received
the certification for our Xtract™ system. This emphasizes once again the distinctive power of the Observation’s proprietary technology.
No other video content analysis solution approaches the high ratios offered by the Observision Xtract™ system. Our performance guarantee
is now confirmed via the i-LIDS certification of the authoritative Home Office Scientific Development Branch in the United Kingdom." 
About the i-LIDS certification of the HOSDB
The HOSDB Library for Intelligent Detection Systems (i-LIDS) is an image library used to test video analytic product performance against an
objective and professional benchmark. Experts at HOSDB conduct yearly testing of video analytics solutions from companies world-wide.
To date, a limited number of video analytics providers have received i-LIDS certification as "primary" detection systems for operational alerts in
sterile zone monitoring applications. The certification as "Primary System" indicates that a detection system can operate as a standalone
system, without the support of a secondary system, due to its superb detection capabilities. 
About Observision’s Video and Audio Analysis Technology
Observision, www.observision.com, utilizes a special video analysis approach. The novelty is that the video analysis function (or application) is
split. In regular video analysis systems a specific analysis function operates in a single pass. This implies that for each analysis function a
video-to-event application is active. Observision utilizes a different method: the video analysis is handled in a two-step fashion: 1. Perform a
complete object analysis on the video stream (real-time), 2, Execute (multiple) rule analysis on the results of (1). The result is a more concise
detection of visible objects plus increased capabilities and performance in handling multiple and also multi-camera rule analysis sets. The Observision
method is labelled as Video Object Analysis. Key is the intermediate generation of the Object Data. As first result this method is especially in the case
of complex, combined and/or multiple rule processing. The Object base method exploits the fact that all objects have been analyzed and stored which yields
that only the rule processing is instantiated multiple times. As a result of Observision Xtract™ system's capabilities, security operators can increase their
awareness of events of interest, while saving labour and significantly reducing the potential for human error.