By Germán Alexis Cortés
However here come two alternatives:
a) That the IP camera behaves like a video webserver . The camera internally has the necessary software to deliver video, manage it and adjust it when from another station on the network, through a browser (Explorer, netscape, opera, firefox), the assigned IP address or name is typed through a DNS server. At that moment the browser presents a web page that resides in the camera, where it is usually started requesting an access code (I must warn that easily hackable) and then the image of the camera is displayed continuously. In this HTML page, there are usually other menus that allow you to configure and make adjustments to the camera. If another user of the network enters the same IP address to access the same camera, it will follow the same procedure until the number of users allowed by the camera is filled. This is a very good solution if we must take the video from the cameras to several users throughout the existing network; however, it becomes another security risk, because precisely anyone just by breaking the access code, can see the information.
(b) Behave like a traditional station; that is, the camera does not have a webserver and in this case if the IP address is typed from a browser, nothing is simply displayed and nothing happens. To be able to view the image on a workstation on the network, it will be necessary to install special software delivered by the camera manufacturer and only through this application will you be able to access the camera, manage it and configure it. In this case, licenses are required for multiple users to access the camera, but from a security point of view it is better, since not everyone has access to the digital video signal. Almost always the software allows to encrypt the information and establishes a secure link between the monitor and the camera.
Which is preferable? It usually depends on the application that is going to be given to the camera. For the case of professional security, I lean more towards the second option.
Power
Another important detail is the power supply of the IP camera. The vast majority of professional equipment has a 12Vdc/24Vac connector, which must be supplied through a centralized source and / or through local sources in each site, - always remember to leave the same electrical phase to avoid problems of synchronism or returns by land-. This involves carrying a power cable of the indicated caliber, which takes into account the distance, output voltage, attenuation and other factors that guarantee the appropriate energy for each camera.
If UTP cable of several pairs is used, this power supply can travel through one of them and behaves according to Ohm's laws. In this case we must be aware that the caliber of the cable (usually 24 AWG) is very small and that surely we will not be able to move away from the centralized power supply. However, everything depends on the consumption of each camera.
If we use UTP cable following EIA/TIA recommendations for structured cabling in the worst case the horizontal wiring will reach a maximum of 90 meters.
There is also the probability of using the new POE (Power over Ethernet) protocol, which allows the power signal to be transmitted through one of the cable pairs and send just over 1 Amp. that exits directly from the active network equipment (switch), which must comply with the standard.
Again, it's one thing what the standards and advances in the IT world do and another what it means for the security world. In this case using PoE may not be highly recommended in electronic security, for several reasons: we are tied to a maximum distance that is not always met. The power comes out of the same active equipment increasing the risk of damage and maintenance cost. The camera must meet the PoE standard and not all do. Finally, it is unsafe to share all data and communications wiring securely.
Beware: here I must warn that some IP cameras claim to comply with PoE, but there are background inconsistencies, such as saying that they are powered at 12Vdc and consume 15W. If a simple analysis is made, it is observed that the current will then be 1.25 Amps and in a 24 AWG gauge cable, it is not advisable to handle this large current. So in this example, you could not use the traditional UTP cable, you will have to buy a larger caliber one... my suggestion is always to do the calculation following Ohm's laws.
Recording of images
A very controversial and unclear issue is where are images recorded from an IP camera? Many people say that in DVRs, however I must clarify that DVRs record non-digital analog video signals. Although many have a network port, there are only signals for other stations, but they are not designed to receive high-speed signals from IP cameras.
It is normal to record the IP cameras using a recording software that is normally supplied by the manufacturer of the camera, on the hard drive of the PC where you are consulting. However, the intention of most software of this type is to record a part of the received video and not to make a continuous or event recording permanently. In this case the software records the video scene in a known format and without many compression and/or security requirements.
When you want to record permanently (that is, 24x365) by events or continuously, it is indicated to use an NVR (Network Video Recorder), which is a PC dedicated exclusively to take the data from the IP cameras through the network port (usually high speed greater than 1 Gbps) and store them on a high-specification hard drive. These computers are programmed once and remain as black boxes on the network, storing all the video. They have the necessary software to differentiate each video stream and store it safely. Manufacturers specify the capacity of an NVR in video channels that they can store simultaneously, the amount of FPS and resolution for each camera.
Recently DVR units have come out on the market that are presented as hybrids. In this case they can record some analog channels on their BNC inputs and other digital video channels through their network communications port.
Be careful not to record the images of your IP camera on the other end of the network, as it would totally occupy it. I recommend that you record images coming from IP cameras on NVRs located on the same LAN and only use WAN channels to monitor some sporadic, non-continuous video signals.
Some recent models of IP cameras allow you to generate more than one video stream . Each stream can be of a different quality and speed, even with a different compression algorithm. This way you could use the good quality JPEG stream at 30 FPS to observe the signal on your own LAN, the high-quality MPEG2 stream at 30 FPS for local recording, and the low-quality MPEG4 stream at 5 FPS, to monitor remotely over the Internet.
Current market
In the current market there are already many cameras with IP outputs, even mobile cameras in high-speed domes. In this case, the video and data for engine control go through the same communication channel. Feeding in these models is recommended to carry separately due to the consumption of the unit.
Some brands, which do not yet have IP cameras, have encoders that allow an analog signal to be passed from any traditional camera to a digital signal, thus converting it into an IP camera. In this case the encoder digitizes the signal, compresses it and delivers it over an ethernet network using a known communications protocol. In this case the IP address is assigned to the encoder and handled as an integrated IP camera.
When to use an IP camera?
The answer involves three clear factors:
1.La security application requires streaming video remotely. Preferably when there are many cameras distributed over a wide territory and you want to monitor and manage from a centralized point.
2.There is already a data network covering the different remote sites.
3.The bandwidth of the data network is sufficient to transmit the video efficiently without interfering with the normal operation of the network.
When factor number one is not met, that is, all the cameras are in the same place, my suggestion for now, is a traditional system, to use analog cameras, although using UTP cable and transceivers. In these cases it is not worth using the existing network infrastructure, because it will always be better for security and administration, to work with a separate network and dedicated exclusively to security. A network that only those in charge of security and specialized technical maintenance operate. Unless you have the budget to build a high-speed data network (remember that at least two to five Mbps are required for each camera), it will always be better to make this traditional type network completely isolated from the data network that carries multiple information.
If factor two is not met, i.e. there is no data network, it is important to know that the cost of the IP camera is minimal compared to the necessary telecommunications infrastructure. Only in cases where: a) the concept of a digital video network is very clear, b) it has been analyzed with respect to other video signal transmission alternatives and c) it is concluded that it is better to use IP cameras; then we proceed to create a completely new data network, with sufficient bandwidth (LAN greater than 1Gbps and WAN greater than 20 Mbps). It is recommended that if a new network is built, it is as far as possible for the exclusive use of video and / or electronic security.
When factors 1) and 2) are given but 3 is NOT met), which is what happens in most cases; it is necessary to rethink the project from two perspectives:
a) update the data network and improve it to achieve the necessary bandwidth or b) sacrifice performance and quality in the images that are sent, obtaining isolated images at very low refresh rates (less than 2-3 FPS) and with resolutions similar to CIF (320x240), in the best of cases. This clearly cannot be used for a security application but for remote monitoring. Don't expect miracles.
The security and systems departments of each company must then reconcile and mutually find an appropriate solution. Many IP cameras can limit their bandwidth at the cost of lowering FPS to very low resolution. With units that have these characteristics, the systems department can assign to security a bandwidth of the network and stay calm, in the sense that the network will not be taken over beyond what is allowed, creating conflicts in the traditional data operation of the existing network. Here the sacrificed is the security department.
In most cases, companies make a significant investment in IP video technology, knowing that they will have images of low resolution and low speed, that is, knowing that the images are going to be deficient. However, they are only aware of their error, when they need to see the image of some sinister and effectively realize that the images are deficient. So what do they do it for? maybe to be fashionable? Why did someone sell them a misconception?
Other companies decide to take advantage of the moment of request for technology originated in the security department to update their data network and increase the bandwidth and sometimes the coverage. However, the cost is so high that many times the budgetary limitation means that the project does not progress.
Common mistakes
•Not realizing that the speed of a LAN is much higher than that of a WAN. So don't expect the moderately decent image you see on the network your office to see from your home or from a remote office.
The quality, brightness, speed and sharpness of images at the analog level does not depend on third parties, it is the camera, the cable and the monitor. The NTSC Standard is met, therefore the image always goes at 30 FPS and with the source resolution of the camera.
•In the digital case, the data communication network depends on third parties (switch, router, other networks, information that is shared with other departments, resolution of monitors, cameras, type of cable, among others); therefore keep in mind that the image is going to be considerably inferior.
•Consider that to view remote images, I must use an IP camera. This can well be done by a DVR recording analog cameras (economical, high quality, professional) locally and sending the images through the network port, when entered from the Internet.
Next step
As we can see, the solution with IP cameras is not the most suitable when external conditions are not favorable. Although not part of the topic of this article, a suggestion to evaluate is to place DVRs, wired locally through UTP and perform sporadic remote access (only when an abnormal situation that requires immediate review) occurs through the existing data network. In this abnormal case, the traditional thing is that it is coordinated with the systems department and priority is given to security so that it can use all the bandwidth of the network for the video signal, temporarily placing the other processes on standby while covering the claim.
When video equipment already exists, an IP system can be involved, through elements that receive the analog video signal, digitize it, compress it and take it out to the network using a known data protocol. In this case encoders from different manufacturers can be used with any type of cameras. In the same way, the equipment to display the images can also be attached to the network through decoders that make the reverse process.
Some major factories currently do not offer IP cameras, but a line of encoders and decoders that transforms the landscape back over IP. The same restrictions and benefits mentioned apply in this case.
The tendency of several companies is precisely to offer a digital solution that gradually replaces the traditional analog solution. The key point is the pace of change. Only until our telecommunications networks are high-speed and excellent performance for permanent video will we come to apply these solutions routinely. For now I believe that we must analyze the aforementioned factors and only when we have the complete certainty of not losing the investment, take the step towards fully digital video. The issue of virtual arrays is nothing more than that. It consists simply of having many video signals coming from many IP cameras or the combination of camera and encoder, in conjunction with video signals to monitors and decoders, network video recorders and workstations that manage to manage the data immediately and efficiently. In this case, the same manufacturers recommend that you work on a very high-speed data network, exclusively for video.
Other topics to be discussed in another opportunity, will be: a) type of software to manage video and give access to the different video units. b) integration with current analog video and integration systems, c) sophistication in video analysis, to support the work of the operator and finally d) security in the data networks that transport video.
* Germán Alexis Cortés is an electronic engineer expert in building automation and General Manager of Insetrón Ltda. It is dedicated to project consulting and specialized training of a Latin American nature. If you wish you can write to the email gcorté[email protected]
Leave your comment