Internet Protocol (IP) Security Cameras

 

The first generation of digital video was justified by satellite TV, since it was not feasible to broadcast many analog TV channels to millions of receivers via satellite.  Signal compression techniques (MPEG), originally developed for this application, are now commonly used in other fields to show video on the Internet or store large time periods of camera footage.

 

Similar digital video techniques are now commonplace in the video security industry.  Unless you buy out-of-date equipment, you are unlikely to deploy a completely analog video security system today.  Digital technology has made pure analog video security technology virtually obsolete.  The digital video recorder (DVR) replaces what would have been a VCR a few years ago. 

 

Almost all systems now include the ability to view the video from your cameras over a computer network, including the Internet.  The difference is where the digital conversion is made, either inside the camera housing, at a cell-controller (near the camera, but outside the camera housing), or at the DVR.

 

The current debate concerns the Internet Protocol (IP) camera.   It is important to note the difference between a “web camera” and an “IP camera”.  A web camera generally connects to the USB port of a computer and utilizes the resources of the computer to function, but an IP camera is independent of a host computer, at least to a degree.

 

 

Are IP Cameras the Future?

 

There is a certain amount of inevitability to the IP camera.  The question is when, not if, the IP camera will come to dominate the security camera market.  IP camera installation is easier because the network cabling is less expensive and much less bulky than the coaxial cable normally used for analog cameras.  There is a certain presumption that anything digital will be better than its analog alternative, but in this case, the reality is more complicated.  Like most new technologies, a certain amount of “oversell” by the IP camera companies marketing their products should be expected.   But if we are to make an intelligent decision, it is important to recognize the issues currently associated with IP cameras.   

 

 

IP Camera Specifications

 

Shop for an IP camera, and you will find cameras advertised as a “1080P Camera”.  These cameras do output a signal using the HDTV standard, but this tells the customer little about the actual camera resolution.  1080P is a display resolution, and an image of any quality, good or bad, can be displayed in the 1080P format.  If one watches an old grainy Charlie Chaplin movie on an HDTV set, the inferior quality source becomes the deciding factor.  For most cameras, IP or otherwise, the deciding factor will be the camera sensor, not the output format. 

 

Similarly, IP camera manufacturers can avoid “apples to apples” comparisons by specifying pixel count.  Pixel count and image quality are independent.  More pixels may lead to better photography, but not necessarily.  A high quality camera will specify all three: sensor pixel count, output format, and tested resolution.    

 

Since IP camera companies often find themselves priced out of many markets, business considerations can force questionable compromises which unfortunately, sometimes lead to inferior cameras. 

 

With cameras (and anything else technical), it can be an indication of a problem when the specification is obscure.  Sometimes the only way to estimate a given camera’s performance is to reverse engineer the camera’s specification.  Find the identity of the camera’s CCD or CMOS sensor, and go to the sensor manufacturer’s website for the datasheet for the module performance.  (“An Engineering Approach to Consistent Images”)  

 

 

 

Problems with Current IP Camera Implementation

 

IP camera technology does have its problems.  The first issue is expense.  IP cameras with comparable image quality are still significantly more expensive than an analog alternative.  The second issue is that the bandwidth needed to communicate the video is a challenge for conventional networks if more than one or two cameras are employed.  When making comparisons between analog and IP cameras, similar image quality specifications should be used, otherwise the decision will based on the novelty of the technology, not the performance.  

 

An IP camera can be found at about the same price as an analog camera, but not if the two cameras have comparable performance.  For example, IP cameras generally have less impressive low-light sensitivity specifications (1-4 Lux typical) than analog cameras (0.001-0.8 Lux typical).  Excellent quality IP cameras are available. The Cisco 2500 has a sensitivity of 0.4 Lux with more than 100 dB dynamic range, but the Cisco 2500 retails for more than $800.   A comparable analog camera would sell for less than $300.

 

The other problem with IP cameras is the signal bandwidth needed to link the camera to the system network.  It is feasible to install an IP camera on an existing network, but the impact multiple cameras will have on your system’s network performance should be considered, especially if many cameras are to be installed.  If a dedicated star network topology (home runs) is to be installed solely for the use of the cameras, then this problem may be avoided (Video Signals over the Network). 

     

There are two approaches to the use of IP cameras.  The first approach converts the video signal from analog to digital at the camera.  This camera will incorporate enough computational ability to function in a network environment, including an operating system stored on a small flash memory module. 

 

The second approach takes the first approach a step further.  This approach digitizes the video, but in a decentralized macro system design.  This camera includes the ability to truly operate independently.  It can store video until needed, and it has true video analytic capability (artificial intelligence).   With this artificial intelligence, if an alarm condition exists, the camera can make this determination, and contact the host.

 

 

IP Alternative 1:  Digitized Video 

 

Digitized video is the more common approach currently.  With this design the digital camera captures the image, and constantly communicates the video as a digitized signal over a computer network to a video server (computer).   The server owns the responsibility for processing and storing the video.  Even though the camera must have some computation capability to communicate over a network, it has little if any on-board capability for logic functions and storage.  Remember that video will require a significant microprocessor, RAM, and video storage for even limited manipulation.

 

 

IP Alternative 2: Decentralized Logic

 

If sophisticated decentralized processing (video analytics) could be cost-effectively deployed on IP cameras, the transition to IP cameras would be dramatically accelerated.  A few companies have introduced digital cameras that have integrated web servers so that no external computers are needed for communication, but this is different from the ability to analyze at the point of capture and store the video locally. 

 

On-board logic and video storage would eliminate the need for all video to be constantly communicated to the host, and would solve the bandwidth problem.  Video that needed to be reviewed could be requested by the host only when needed.  

 

There are cameras available with these features, but they can be very expensive.  Sometimes these functions are added in a separate module local to the camera called an “Intelligent Video Server”. An example of a design with logic and storage are some of the cameras offered by the Swedish company Axis Communications (Note:  not all cameras from Axis are suitable for a decentralized logic system).  

 

When large systems of IP cameras are employed, if a dedicated star topology network is not employed, compromises will need to be made to reduce costs and bandwidth requirements.  The resolution and frame-rate are the most common compromises. 

 

A survey of technology professionals estimates that the IP camera will become dominant when there is less than $100 difference from the analog camera.  As the price of technology continues to drop, the IP camera will be more competitive, but it could take many years. 

 

 

About the Author:  Brian Bradshaw is General Manager of B.V. Technology.  He is an InfoComm Certified Technology Specialist (CTS), a Microsoft Certified Technology Specialist (MCTS), and CompTIA A+ computer technologist. 

Website:  B.V. Technology

 

 

Reference