Smarter cameras lead IP video evolution

Application integration shouldn’t be a pricey add-on

There is a lot of misinformation about IP video systems that comes through in many of the articles written in print and on the Net. It appears that many people interviewed on the subject really do not have a grasp of it or are not actively involved with IP surveillance on a daily basis, and while they give their opinion, a lot of it is not really based on what is actually happening and why.

IP surveillance is soon to enter what financiers refer to as a "bull market". It hasn't happened yet, but it will. When I say bull market, of course I don't mean the prices are going up, but the sale of IP cameras is going to absolutely explode. This is newsworthy. But what is more interesting is why this is happening and what is creating the demand. A lot of articles are written about the virtues of IP cameras, such as the fact they plug directly into the network allowing users to leverage existing network infrastructure, or that it's "future proof". These are really peripheral reasons why the market is shifting toward IP-based surveillance systems.

The reality is, end users buy IP camera systems for their superior image quality. Customers are upgrading their current systems in favour of a superior system that provides usable video. Current DVR (analogue to digital systems) are restricted by PAL, an analogue protocol, which can only offer a maximum of 288 vertical lines of usable video, when there is motion. Recording in higher PAL resolutions, such as 576 TV lines, causes "image tearing" (ripping or distortion in the image) when movement occurs. This kind of low-class resolution prevents facial or object identification, especially over distances and when trying to extract clear stills.

Better resolution explains the worldwide trend toward IP surveillance systems. Think about it, if IP-based cameras could not deliver better resolution than a DVR surveillance system, companies would not fork out money to make the transition - where would the commercial viability be in doing that?

While the demand for high-resolution video is the driving force behind sales of IP camera systems, this also creates some level of pain for the customer buying the system. For example, an educational institution recently purchased just over 70 IP cameras to replace an existing DVR surveillance system.

These new IP cameras were set up to record in VGA resolution (640x480 pixels), using MPEG-4 compression, at five frames per second and to speed up on motion to 15 frames per second. Since the cameras were recording in MPEG-4 compression, the customer also had to also install a server that would reconstruct (decode), the compression to a usable video format. This meant multiple Xeon processors being specified into the recording server to cope with the decoding, plus SCSI disks to cope with the write speeds required. The server cost around $25,000.

On top of that the customer had to buy "pay per licence" video management software, which had to be installed on the recording server to enable management of the IP cameras, which was another $15,000, plus an annual maintenance fee for software upgrades of $2500.

Imagine if the organization had installed megapixel IP cameras? They would have needed three of these servers just to cope with the decoding!

This problem is the same for every single brand of IP camera on the market using MPEG-4 codecs.

Smart IP cameras are now available that can actually decode the video inside the camera first, so when the video leaves the camera it's already reconstructed. This means that the bandwidth usage is very lean and the cameras can record directly to a network attached storage devices. Now there is no need for expensive servers with multiple processors and high-speed drives with the latest evolution of IP cameras. This equates to a massive cost saving to the customer.

NAS devices are a lot more reliable for recording video, cost less per gigabyte and there are no software licensing costs to factor in.

Such cameras can also send the megapixel video in a single stream. Typically, MPEG-4 cameras have to send multiple streams to achieve high-definition resolution!

So this now leads us to application integration. When you talk about integration with other devices using IP camera systems, it sounds easy on the surface and all the IP camera manufacturers will tell you their technology can do it. And they can, but not without the pain of cost, as a lot of links have to be added to the chain to create the final solution.

The key to successful integration lies within the software architecture of the IP camera system. However you will find almost all of the IP cameras on the market rely on third-party management software to enable this sort of integration - which again equates to additional cost.

These software companies are taking advantage of, and cashing in on, the limitations found within most IP cameras. There are software products out there which enable, access control or POS integration, but these are not cheap solutions, and still require a certain level of coding for final integration.

To avoid this, look for video management architecture that is built into the camera. This should be free and enable fast integration to all sorts of IP and serial devices like access control and POS systems.

Simon Hall is managing director of The IP Academy, a channel marketing company specializing in IP camera and network video technologies. The IP Academy covers company news for MOBOTIX IP cameras.

Join the newsletter!

Error: Please check your email address.

More about GigabyteMotionPLUSSpeed

Show Comments
[]