Disqus - Latest Comments for dougmarman

Re: Fike SigniFire Examined

Doug Marman — Wed, 28 Mar 2012 15:10:50 -0000

John, there is a problem using traditional spot-type smoke detectors in large warehouses and atriums: The stratification of smoke. Air layers prevent the smoke from reaching the ceilings. That's why you will usually find beam-type smoke detectors in those applications, which shoot infrared beams of light across the space. The problem with beam-type smoke detectors is that they detect only a small linear beam area, so you need quite a few to cover an area. They are also quite a bit more expensive and more time consuming to install than spot-type detectors.

I believe video-based smoke detectors will be far more in competition with beam type smoke detectors than spot-type smokes. They will also provide better coverage of a large area.

The biggest issue in video smoke detectors is agency listing. Fire authorities, who decide whether the proper fire alarm equipment is installed to meet codes, require all fire products to be either listed by UL or Factory Mutual, or another recognized independent organization who has tested the equipment. The Fike product (originally Axonx) was the first to have such listing, from what I know.

Re: DVTel to Merge & 'Reverse IPO'

Doug Marman — Mon, 07 Nov 2011 12:49:57 -0000

John, yes, there was a large example of reverse merger: Interlogix was formed by the reverse merger between SLC Corporation (Security & LifeSafety Corp) and ITI (Interactive Technology Inc.). Interlogix was sold to GE within a couple years. It had revenues of about $800M when it sold to GE.

Re: Testing: Genetec's Trickling in Omnicast 4.7

Doug Marman — Tue, 15 Mar 2011 19:40:26 -0000

First, to help clear up some confusion, solid state memory is not limited by the number of times you record to the card, but the number of times cells of memory are rewritten. If you are recording continuously to the SD card, as long as you continue to record to new sectors, it is considered one rewrite. If the camera is set up to use all the sectors of the SD card, then after the card has been fully recorded, it would start over again. That would be a single rewrite.

If someone has enough memory in their card to hold 3 days worth of video, as John was suggesting, then that means it will only need to be rewritten once every 3 days, if you are recording continuously.

Unfortunately, the lifespan for most SD card is not 100,000, but 10,000. That's because the most popular and least expensive cards use MLC type memory, which has a shorter life than SLC memory cards (which are more expensive and harder to find.)

The good news is that this is still adequate. 10,000 rewrites every three days is over 82 years. I've spoken to some engineers who have been recording video onto SD cards continuously for over 10 years without any problems.

The second thing I wanted to mention was in response to John's comment that Milestone has no cameras with built-in storage in production that they support. Not true. VideoIQ's cameras are supported and have been for almost a year. VideoIQ's edge storage is also supported by IPVision Software video management software. We hope to add support for Genetec's new trickling capability soon. We think it is a great addition. We of course also full support through the VMS that comes bundled with our cameras as well.

Re: Product Sales Harm Security Users

Doug Marman — Sun, 20 Feb 2011 21:52:34 -0000

I don't disagree. You would think it is simple enough. But don't underestimate how even the simplest things are over the heads of most people, unless they have some training. Most people could learn it and figure it out, but don't have the patience to do so.

It isn't that it is that difficult to learn, but without someone guiding them, they don't know what they don't know, so most are reluctant to do it on their own. This is just the nature of technology designed for professionals, rather than consumer electronics. Even some consumer interfaces are too complicated.

This little issue might seem insignificant, but it changes the whole nature of how such products need to be distributed and sold.

I agree with you that this is one reason analog is here to stay for many years.

The strange thing is that many users will probably have less problem with the IP part of the installation than many of the security video oldtimers, but the users will not know much about camera placement, lenses, proper lighting, at other basics. They'll aim a camera at a doorway entrance and complain that the picture is all washed out. I'm sure you know what I mean.

Re: Product Sales Harm Security Users

Doug Marman — Thu, 17 Feb 2011 15:21:48 -0000

The video surveillance marketplace is a complex system. There are many factors driving the way the market is today. Looking at just one piece of it, in this case the sale of equipment, misses the real causes that have led to the system we have.

When products are simple to understand and select, then direct sales work well and are the best approach. But surveillance systems are too complex for this, in most cases. So, you need to add expertise somewhere in the loop to make sure the right products are being selected for an effective solution. That's why integrators exist and are not going to disappear anytime soon.

If integrators are involved in helping the end user choose the best products to provide the security protection they need, then why shouldn't they be paid for doing that? If integrators are going to be servicing the system, then they also want to choose products that will work reliably and stick with something they know. It is a lot less expensive to manage familiar products. That's good for them and the end users, and puts pressure on the manufacturers to make reliable products.

Manufacturers don't have the ability to get the product expertise to the end users directly. If you could solve that problem, then the system would change completely. But, I agree with Ksenia, I don't see how that happens.

Re: Training: Remote Video Surveillance Monitoring

Doug Marman — Mon, 26 Jul 2010 18:46:21 -0000

That's exactly the problem as I see it.

If you, or others who are objective, can test systems and let others know which ones are the few that "work", it would be a huge benefit to the industry.

Other than that, word of mouth and testing it yourself are the only solutions I know.

Re: Training: Remote Video Surveillance Monitoring

Doug Marman — Mon, 26 Jul 2010 13:25:35 -0000

I agree. There are only about 3-4 globally that will work well in outdoor environments without producing a lot of false alarms. And some of them do require specialized training to set up properly.

I didn't think 3-4 sources represented a shortage, since most people will use only one product line for a job, or in very large complex jobs might want two product lines to choose from. So, if they know which 3-4 really work, that would be good enough.

Perhaps I'm missing your point. Why is it important to have more sources than this? I know it helps to have options to choose from, but isn't the real problem that it becomes critical to know which ones to use when there are so many that don't work well? Or were there other reasons you were concerned about?

Re: Training: Remote Video Surveillance Monitoring

Doug Marman — Sat, 24 Jul 2010 18:21:43 -0000

John,

Is it a good idea for me to name names in public? I'm reluctant to do that.

On the other hand, I'm concerned about the exact issue you are raising: There are way too many poor performing analytics, which are causing grief for integrators and end users.

I'm especially blown away by the major brand name camera companies who are selling and promoting solutions that aren't even close. We've seen some of the worst problems in the field from these companies who sell advanced motion detection for outdoor applications and end up having to rip the systems out. This is seriously hurting their reputations. How is that a wise thing for them to do? Or are they really this clueless?

If you want to talk on the phone, I'd be glad to offer some guidance on which technologies we see as best. Then, perhaps you can test them and verify.

My best recommendation for integrators and end users at this point is to be sure to run a test first. Don't take any manufacturer's word. Test against our VideoIQ systems, if you want a reference to compare against. We'll be glad to help with that. A lot of our products have been used to replace other systems that fail, but this is a painful experience for integrators and end users. Those who have avoided the problems by testing first have been much happier.

Doug.

Re: Training: Remote Video Surveillance Monitoring

Doug Marman — Tue, 20 Jul 2010 17:41:19 -0000

John, nice overview.

In support of your claim that video analytics can provide a major benefit with remote monitoring, I asked some monitoring companies who monitor auto dealerships to compare their results using traditional approaches (using outdoor motion and fence sensors with video cameras for verification), compared with systems using video analytics to trigger alerts.

They could monitor about 10X as many sites per operator when analytics sent them alerts with video clips of the detected event.

Even more importantly, their protection improved dramatically, since they were less likely to miss alarms.

Another thing worth adding is the value of having audio over IP available. This gives the monitoring personnel the ability to respond immediately, which means being able to stop crime or break-ins before they happen, rather than just recording it. That provides a much greater level of protection for the money spent on security.

I don't agree with you that the problem with video analytics is the shortage of systems good enough to provide quality detection with low false alarm rates. I think the real problem you are getting at is that there are way too many products that are promoted in such a way that you don't know where to use them reliably.

There are many brand name companies selling systems that will rarely if ever work reliably outdoors, without making those limitations clear. Other systems might work fine with people counting, but not perimeter protection.

This problem can't be solved 100 percent, but if manufacturers would take the responsibility to clearly advise integrators where to use and not to use their systems, this would make it much easier to find the products that will work for the needed application.

Re: Do Video Analytic Providers Have to Sell Products?

Doug Marman — Tue, 06 Apr 2010 18:59:30 -0000

Yes, we are finding wide interest across lots of markets. It didn't make sense for us to go after one vertical. But with a software only solution, it would have made more sense.

However, as you point out, verticals are not easy either. You need to be an expert on the vertical market you are going after. Retail has lots of potential long term, but it is probably the slowest market to grow. Many retailers are just adding video. They are not generally your early adopters. And the consumer market is even tougher if you aren't an expert in that area.

Re: Do Video Analytic Providers Have to Sell Products?

Doug Marman — Mon, 05 Apr 2010 20:03:38 -0000

John,

I think the problem is more challenging than it sounds. Let's say you try to keep your overhead low, just have a few engineers working to develop something. You have no sales force, no marketing dollars. You just go around to the camera or DVR manufacturers to sell them on your technology.

The first thing you discover is that it takes 2-3 years, after you've finished developing your first release, for testing and installing your systems in the real world to discover all the problems that looked so much simpler in your lab. This is the biggest mistake everyone makes. I haven't seen anyone do it in much less time than this. So, you've now spent millions of dollars.

Then after you've gone through all of that, and if you are lucky enough to license your first customer, now it is going to take 6-12 months before their first product ships. It could be a lot longer. And your future totally depends on the success of that product. If the manufacturer did a poor job, it isn't going to sell well and it will reflect poorly on you.

More significantly, if they don't put a big marketing push behind the new product, like you would do if it was your product, then it might not get any attention at all, even if it worked perfectly.

In other words, you are not in control of your destiny.

This is very risky in a new market.

Then, probably the biggest problem of all is that if you could develop such a system with just a few engineers over 3 years, then you can't expect to be in business long before all the camera and DVR manufacturers develop it themselves. Especially if they think it is going to be important and will be widely required in future systems. Or if they do continue buying it from you, they are going to keep asking for lower and lower prices because they've got a dozen other small groups who have something that would work and who are selling for less.

You need either significant intellectual property and something very unique and much better than anyone else, or it isn't worth spending all of that time to get squeezed out like that. And you need to expand fast enough in the market or the big established players will just copy you and put you out of business.

The other possible way you can succeed is if everyone knows it is not a big market opportunity, so very few are willing to go after it. Then you won't have a lot of competition and the big manufacturers would be glad to license something rather than spending their own resources on something that isn't very important.

I think it is much wiser to go after a vertical market and try to develop a full customized solution. We've seen companies try that for retail, one of the biggest vertical markets. There are some companies trying this in transportation. It's not easy, and you need to know the vertical market well, but it is less risky than licensing. If you start early enough before the market develops, you can be there when it starts to take off.

In some industries, like networking, you can go out and sell one or two large manufacturers on a new software development and have high revenue sales in short periods of time. That isn't possible in the surveillance business. No DVR or camera manufacturer has enough market share.

Just some more things to keep in mind.

Re: Why Education is Not Enough for Video Analytics

Doug Marman — Thu, 01 Apr 2010 22:05:23 -0000

John,

What you are saying is only true if the technologies aren't good enough. That's when education isn't enough.

Take the early days of wireless alarm systems. There were quite a few products that weren't good enough. Dealers tried to use them, because the benefits of wireless were significant, especially in certain applications. But they would soon give up on them, and word spread about their problems.

Other technologies showed promise. Worked well enough, even though the dealers needed to learn all kinds of special tricks and hints on how to get the most reliable systems. These products sold and the few companies who did this well grew fast.

They spent their time on training dealers and developing new tools. Education was enough for them. Making their limitations clear didn't drive dealers away, but made them feel more comfortable that they knew where to use it. This came with experience and training. The value was high enough that those few that were good enough grew fast.

The early days of passive infrared sensors went through a similar curve.

Unfortunately, I agree with you that a number of video analytics technologies are not good enough. But this doesn't mean that all are. If you can get good reliable systems installed on a regular basis, then education is something that helps accelerate sales and satisfaction both together.

Re: Archerfish Analytics Installation Restrictions

Doug Marman — Thu, 01 Apr 2010 12:42:33 -0000

John,

Like you say, because they are trying to sell a very low cost solution, they are probably using a lower cost imager, which usually requires about 10X as much light. Those imagers work fine indoors, but most of the applications for Archerfish are probably outdoors.

Things like this often mean the technology is not mature enough for an application. I think the consumer market is one of the toughest to make a success in with new technology. The product needs to be super simple to use and priced very low to capture widespread interest.

If professional installers don't find is easy to use, then consumers are not going to be able to use it.

As an example, X-10 generated widespread sales to the consumer market when they got their wireless cameras down to around $100. But they had a 35% return rate because the wireless often took fiddling with to give good quality video. That kind of return rate ruins any chance of success.

The wireless technology was simple enough that most alarm installers know how to make it work, but they've been through years of training. It needs to be a lot simpler than that for consumers to use it widely.

I think one day there will be a big market for video analytics in the consumer market, but my guess is that it will be widespread in the professional surveillance and alarm markets first.

Re: Can Video Analytics Stop Suicide?

Doug Marman — Tue, 12 Jan 2010 16:59:03 -0000

I would want to know more about the previous cases. Are they jumping in front of the train just before the train arrives?

That would be my guess. If that's the case, then what difference would an alert make, if there wasn't enough time to respond?

Even if you could trigger an alert to the train conductor, and the brakes could be put on immediately, could it stop in time? If not, I don't see how this helps.

However, if people are jumping out in front of the train with enough time to spare, then why not look for a person on the tracks? No one should be on the tracks. You should be able to make this work without too many false alarms.

Doug.

Re: AgentVi Says They Can Prevent Airport Terminal Shutdowns

Doug Marman — Tue, 12 Jan 2010 16:31:14 -0000

John,

Yes, a 30 degree down-tilt is fine for us. We work well from 5 - 45 degree down tilts. So, 30 degrees is perfect.

The crowding is something to be more careful about. I watched the video and in most cases the traffic was light, but occasionally a crowd of people passed through. It looked like a reasonable application.

We've seen other systems trying to do direction of travel that have been taken out because they had trouble with this, but ours were put into those same locations and worked fine. However, if it is crowded all the time, I'd recommend testing first. In fact, testing is a good idea anyway. We'd be glad to loan one of our analytics cameras for testing.

Having a 30 degree downward angle helps a lot, since you can see across the people much easier. I don't know how you could do this without something like a 30 degree down tilt.

Doug.

Re: AgentVi Says They Can Prevent Airport Terminal Shutdowns

Doug Marman — Tue, 12 Jan 2010 13:02:53 -0000

We've got airports using direction of travel alarms and they seem to be happy. It saves them a lot of money from having to have a security person sitting watching an exit. Also, we've seen video of the actual event at Newark airport captured by a nearby camera. Direction of travel should work for that application, as Ethan says.

If they could alert a local security person who could get there right away, then they might not need a guard there all the time.

However, if the person gets by that area and into the terminals, it is unclear how much better a video analytics system would be able to track someone down. It might not be much better than looking at video manually.

Our technology (VideoIQ) has the added ability to search by appearance. So our video analytics would have an advantage in this application. But in crowded areas like airport terminals, where there is a lot of obscuration going on, and people walking in groups, it will only help the process. It won't be something that will work all the time.

In this particular Newark case I think it might have tracked the man once he was in the terminal, since he was wearing a long tan overcoat that was different from what most people were wearing. It stood out even in the grainy poor quality video.

I think these kinds of applications are worth testing and trying out. Video analytics could end up being very helpful. But until we have a large scale deployment to try tracking a person through the terminal, I will have to reserve judgment on how effective it will be. I think it has promise, but should be treated as experimental in crowded environments like that.

If Agent-VI had positioned this as a potential way of solving the problem, and something that should be explored and tested, then it would be raising awareness on an important opportunity.

Perhaps some marketing people got a bit carried away. I agree with John that it is not good to oversell and make it sound like something that is easy to do.

But I do think this as an opportunity worth testing, since the cost of having to evacuate a terminal is incredibly expensive, not to mention the huge impact on thousands of people and delayed flights throughout the country.

Doug Marman.
VideoIQ

Re: What Video Surveillance should a Corporation choose?

Doug Marman — Tue, 10 Nov 2009 11:30:44 -0000

John,

Using video analytics makes a lot of sense, since they are monitoring alarms. They will get far more value from their video if it can alert them to potential threats.

However, I would rank perimeter protection a lot higher than you do. Of course, it does vary based on the business. Some don't care if someone is jumping the fence after-hours. But without a doubt, the best way to improve protection is to get better site awareness and early warning.

Perimeter protection has been expensive and prone to false alarms in the past, so you needed video anyway to verify. But with good quality video analytics, the false alarm rate is low and it is very easy to verify by looking at a video clip of the event. The colored boxes show you exactly what caused the alarm, so it takes usually less than 15 seconds to verify.

Loitering, with a dwell time, is a good way of reducing false alarms in the parking lot.

Another thing I would add is audio talk-down, so that you can immediately interrupt the problem before it happens. If someone breaches a perimeter, a warning that they are on private property, that they have been recorded on video and that the police will be summoned if they don't leave immediately gets rid of 99% of the intruders before they can do any damage. With a parking lot, you can make the message more friendly, saying that the lot is protected for visitor security, if you need help finding where you vehicle is parked, come into the office. We can assist you.

They get the message and leave if they were looking for a car to steal or vandalize.

If some of the locations are very small, where only a few cameras are needed, then storage in the camera can be a savings, rather than putting a DVR at the site.

These test cases are valuable exercises.

Thanks.

Re: What Video Surveillance Should a Small School Use?

Doug Marman — Thu, 29 Oct 2009 22:49:34 -0000

John,

There are Remote Guarding companies that will do the monitoring for them. The cost will vary depending on what they want. Monitoring includes the talk-down response to the site within seconds of the event, warning off the intruders, and police dispatch if they ignore the warning. Everything is recorded at the central station, in case a review of the events are ever needed.

Can they hit your monthly price target? I can't speak for them on pricing, but I think some would be in the ballpark.

Many of the companies doing this are quite mature. We recommend our Remote Guarding partners all the time and have talked to lots of happy customers.

I like your suggestion and think it would make sense is if one of the Remote Guarding companies started offering a solution specifically for schools. There are a lot of small independent schools around, but they often talk with each other and recommend solutions. But monitoring centers don't usually have sales feet on the street to get the word out. That's where integrators come in.

The monitoring companies will gladly work with integrators, and of course we will gladly sell them our equipment, so they can offer a turnkey package to schools.

Call me if you need any help with this.

Doug.

Re: What Video Surveillance Should a Small School Use?

Doug Marman — Thu, 29 Oct 2009 20:19:24 -0000

John,

No question about the cost difference. That's why I said you could only fit about three of our cameras into that budget. I don't know enough about whether that solves their problem or not. But it sounds as if they could stop the vandalism problem they are having, which sounds like their goal. They might only need one or two cameras, if it is only some air conditioning equipment they want to protect.

I'm only suggesting this because it could really solve their problem. Recording the video to look it later doesn't stop the crime before the damage happens. It may not even be enough to catch those who did it later.

End users get this difference. Talk to them about it. Run it by them and see what they say. They know that stopping crime is far more important to them than just recording it.

I'm not sure those Bosch cameras are the best choice. Bosch makes some nice cameras, but the one you are looking at is not designed for outdoors, unless it is in an extremely well lit area at night. It has a 1/4" CMOS sensor, with only 1 lux rating and no day/night filter. It looks like an indoor camera to me, with those specs. Plus it only has video motion detection, which as you pointed out will not be adequate for outdoor detection, so you can't get any useful alerts from it. The false alarms would drive them crazy and make it unusable for proactive detection.

So, with the Bosch system, you can afford to put up more cameras in that $6,000 budget - maybe 8, which is more than 3 of ours - however, it will only record the event so you can look at it later. And because it is only a VGA camera, you probably won't be able to identify who it is well enough to win in court. If it is a poorly lit area, you will need to add additional lighting or pick an outdoor camera, which will cost more.

Run the option by your school and let them tell you what they think, rather than ruling it out.

It doesn't work for everyone, but outdoor assets and perimeter protection are perfect applications for video analytics and proactive protection.

We'll be expanding our line soon, and more VMS integrations are coming along.

Just wanted to offer another suggestion.

Doug.

Re: What Video Surveillance Should a Small School Use?

Doug Marman — Thu, 29 Oct 2009 18:33:19 -0000

John,

As you know, our cameras are much more than $500, but they do provide on-board storage that is very easy to access, and include true video analytics that work reliably outdoors. The software that comes with it allows users to do their own monitoring from a PC, or the cameras can send emails during events.

With the audio connection, they can talk down to the site from their PC to tell intruders to leave, stopping the vandalism before it happens.

You can easily set up the rules to only send alarms after-hours or on weekends.

You won't be able to put up 10 cameras for $6,000, but you could put up a few and target the problems they are having and offer a real solution that can stop the crime before it happens.

With the software that comes with our camera, they would have an easy to use monitoring station, or they can access alarm events via a browser.

The other advantage of a distributed system, with storage in the camera, is you can start with one or two cameras and add more when you need them. No need to put in a full 10 camera system if it isn't needed yet. Add a camera when and where you need it.

We've had quite a few schools using our cameras for exactly this case.

Campus Safety Magazine recently published such a story. A school that was having windows broken. The problem was getting worse every year and finally became so bad they had to do something. Video systems alone were solving the problem. With video analytics and monitoring the alarms, the problem was solved. It helped reduce the sense of "chaos" that came with the crime as well, and made the school a better place for the students to focus on learning.

See more about school applications at: http://www.videoiq.com/solu...

It might be a good fit.

Doug Marman
CTO
VideoIQ

Re: Riskiest Companies 2009 - BRS Labs and TimeSight

Doug Marman — Tue, 15 Sep 2009 15:55:13 -0000

I can comment on the first point by saying that our technology has no problem detecting people through chain-link or other fences.

Here is a link to our videos: http://www.videoiq.com/prod...

If you look at the first one, you will see a series of videos capturing people as they walk behind fences. Chain-link fences are actually easier than the fences we show. It also shows how our analytics learns the environment within minutes, and continues to learn and get smarter.

As you say, John, the movement of trees is regularly ignored by most good video analytics technologies.

And no tripwire is needed. As an example, with our iCVR, you just plug it in, aim it and focus it. No calibration or tuning is needed, and it is already configured to detect people or vehicles entering the scene.

Nor does the distance to the person from the camera seem to be farther than the limits of most good analytics systems.

So, the claim that no other analytics can handle the complex elements in the scene is not true.

Doug Marman.
CTO
VideoIQ

Re: What's the Future of SD Cards for Video Surveillance Storage?

Doug Marman — Wed, 19 Aug 2009 22:26:46 -0000

Another great topic, and good discussion.

John, as you know, I've been studying this topic for a number of years. I had some thoughts to add, but didn't want to take up too much space here, so I posted my comments on my blog:

http://spotonsecurity.com/2...

I wanted to add a few more bits of data here:

The life expectancy of SSDs are higher than SD cards. The first reason for this comes down to the wear leveling, error correction and bit management that goes into the SSDs. SSDs do a better job here. But the most important factor is how often are you erasing and writing to each cell of memory.

The most popular SD cards use multi-layer memory cells that have a lifespan of 10,000 write/erase cycles. Many SSDs are using single layer memory that has 100,000 write/erase cycles. However, if you have enough storage in an SD card for one week or more, you should still have a life expectancy of over a hundred years. Larger SD cards should be even better. But this could be one problem with using an 8 GB card and storing to network server.

CameraMan is right that there is a benefit of video sent as a file transfer than streaming video. The IT managers do like it better, because it doesn't demand a given amount of bandwidth, but can resiliently adapt to the bandwidth available.

There is also another factor is is often overlooked. The most common video streaming protocol is UDP, but it doesn't guarantee data accuracy. If data gets corrupted, it just keeps on sending the next frames. TCP file transfers will go back and get the packet over again if it was corrupted. Multicast streaming protocols also have data loss. You can send streaming video as if it were a file to overcome this problem, but this slows down the video stream and requires even more bandwidth.

However, storing all the video onto a server won't save bandwidth. I agree with John on this. It does add some interesting advantages, but i like full storage in the camera better.

Doug.

Re: Are Public CCTV Cameras an Utter Fiasco?

Doug Marman — Wed, 22 Jul 2009 14:29:23 -0000

John,

I agree that jumping in to retrofit an existing city system today is premature.

I've been talking with groups about getting more real-world test results to show the improved performance. I think that the next best step is running tests of Remote Guarding systems, where the analytics trigger alerts with video clips, monitored by people who can then respond immediately via IP audio, thus interrupting the crime.

Remote Guarding has proven itself in many commercial applications, delivering true crime prevention that works, and there is good reason to think that it will work just as effectively in city surveillance.

But I think the next step is setting up systems and carefully measuring the results in a scientific way, so we can see how much better it is.

Once the value equation is proven, then you can estimate more accurately what the return on investment will be. Then you can justify retrofitting existing systems and show how quickly it will pay back.

If you look at the commercial applications where Remote Guarding has been used, we've seen cost reductions from 50% to 75% compared to on-site guards, while reducing crime significantly at the same time. So, the potential exists to make a police force far more effective, and significantly impacting crime.

This could be a valuable change in our approach to crime reduction. But we need some more concrete tests run to evaluate exactly how effective it is. Then you can justify the expenses and the efforts to retrofit or add new systems.

I think all of the technology complexities are workable.

I do agree that tracking logos on clothing is not one of the areas I would focus on at this stage.

Re: Top 5 IP Camera Problems - 2009

Doug Marman — Fri, 17 Jul 2009 15:54:35 -0000

The implementation of our analytics in our iCVR is simply another generational improvement over what we've been using for years. It is hardly fair to say it is not mature. I'm sure IOimage and Agent VI are continually making improvements as well.

When it comes to performance, in this case I know we have for more experience than you do, and we have long list of companies who have tested our products against the others you have mentioned, along with many others.

These are not grand statements, but actual real world results.

Re: Top 5 IP Camera Problems - 2009

Doug Marman — Fri, 17 Jul 2009 14:56:25 -0000

None of what you just wrote makes sense. I thought you said you were well informed.

Both Agent VI and IOimage would be more expensive, both from an equipment standpoint, and also from the installation cost. Both costs would be more expensive.

Remember, Malmo, Sweden is in Axis' backyard, and they work with Agent VI. There is no way that solution would have been less expensive. Not even close. Plus far more complicated.

Both IOimage and Agent VI require calibration of every camera. Integrators tell us that it takes at least 3 hours per camera and more often quite a bit longer to calibrate and tune each camera. Our iCVR needs no calibration and tuning.

As for mature, that doesn't make sense either. The first VideoIQ products were introduced in 2003. I don't know if IOimage was working with the Israeli military before then, but if so, not by much, and certainly Agent VI was not even in existence back then.

The first embedded products using VideoIQ technology were introduced in late 2005 and sold tens of thousands of units into Europe. I believe it was one of the first embedded video analytics products and was certainly one of the most successful so far.