Over the past few years, a huge number of network cameras have been introduced to the market and installed in various projects. Trade journals devote themselves to the topic with fascination, test such cameras or, at least, publish an overview of currently available products. No matter how laudable the intention of providing customers with a general view may be, a significant amount of irritation remains with the customer. This is, among other reasons, due to a sometimes overdrawn and confusing description of fundamental issues under the subject of "Video over IP". Even pointed inquiries are often not helpful when trying to figure out whether or not a product is suited for professional security applications.
In answering this question, it is not just about individual components of the installation but the complete system. Can complex visions with an extensive networking concept actually be realised? The limitations of a system become quickly apparent once a non-uniform transmission of motion sequences is visibly emerging or when system failures become the rule rather than the exception. Is there a suitable video management system in place which allows the user fast and efficient access to the recorded video material? Can the recordings of various cameras be archived for longer periods of time and replayed synchronously? Is the image quality sufficient and is the system operating reliably? Those aspects determine the value of the entire system. This shows very plainly that the term "Video over IP" should be expounded from a user perspective, whereby the role of analogue and IP cameras within this concept is absolutely in need of explanation.
Video over IP with analogue and IP cameras
The network technology allows the user to install various recorders and IP cameras throughout a building and all across the premises, enabling a high degree of flexibility during the project planning for a CCTV system. In order to make ideally profit from the benefits of an IP system, the video management system needs to be designed accordingly.
The video management system should allow for the fast and efficient search for recorded video data, live images and further data provided by integrated third-party products. It does not matter where the recorders are installed or how many there are within the system. Moreover, the video management system is able to access the images regardless of the data format, e.g. wavelet, MPEG-2/-4 or H.264, as well as of the number of cameras used by the pool of recorders.
So as to facilitate the access to the live image as well as the recording, all recorders are addressable via their IP identification. Within the pool of IP addresses, IP cameras that are integrated into the system are directed 'next to' the recorders. Analogue cameras, however, are directed to the channel 'behind' the recorder, which is accessed via its IP address. Therefore, as far as the use of the image material is concerned, it is irrelevant whether an analogue or IP camera is activated and viewed on the monitor because the video management system is able is able to display video images of any coding standard.
There are some significant differences between the remote control of analogue cameras on the one hand and network cameras. An IP camera is configured via the website it provides on the network, whereas analogue cameras become configurable through the management software. When using the software, the recorder to which a certain analogue camera records is activated. Analogue cameras with Cam_inPIX technology are remote-controlled directly from the recorder via the video cable (coax). This unique concept is called 'UTC' control (Up The Coax) as no further cabling besides the video cable is necessary to configure the camera. The remote control of the whole range of functions of PTZ dome cameras can be done via the interface of the recorder using the user interfaces of either the management software. The management system allows for the control of analogue cameras, providing the user with the same controlling capabilities as is the case with IP cameras.
The major difference between the remote control of analogue cameras on the one hand and network cameras on the other hand is the type of connection to the recorder. Analogue cameras are recorded via the video cable (coax or UTP - Unshielded Twisted Pair) instead of the network. Therefore, the recording of analogue signals is always guaranteed, even if the IP network is overloaded, structurally flawed or temporarily malfunctioning. The IP network thus has to be designed with sound judgement according to the performance requirements and considering possible disturbing factors.
Lack of standard
Discussions relating to IP cameras are often dominated by only a few individual features and in analogy to the area of photography the trend towards megapixel resolution is often evoked. What may be a huge benefit for photographers who usually work with mostly prearranged, individual images, can turn into a curse when it comes to the transmission of video images over a network.
An exceedingly high resolution of the camera sensor takes up significant network resources and consequently a lot of storage space with higher frame rates. Megapixel cameras are usually a sensible choice when it comes to monitoring large areas with only one or two cameras, because then it becomes essential to have a wide zoom available for individual areas of an image. If the camera's field of vision is limited by obstacles such as walls or high shelves, even the use of megapixel cameras often makes no sense.
Moreover, there are a number of ways to enhance the richness of detail of a video images. The innovative Progressive Scan sensor technology Cam_inPIX, developed by Dallmeier, allows for the reliable reproduction of more details in shadowed as well as bright areas of an image. This is possible thanks to the Ultra Wide Dynamic Range (UWDR) feature and the high colour differentiation. The excellent regulating behaviour of the camera module is a decisive advantage at this. In this context, simply put, the motto should be: "The proof of the pudding is in the eating." Testing any camera 'A' against any camera `Bˇ allows for a comparison of the quality of the recorded images. With relevant scenes degree of detail recognition should be compared to the storage space that is actually used up.
It is, however, difficult to find out the performance figures relevant for the video stream. The data sheets of network cameras provide the maximum data rates at which video images can be sent via the network. Only users with a handful of IP cameras in their networks could come up with the idea of trying out that setting. No camera manufacturer, however, states the minimum data rate necessary in order to ensure a high image quality when transmitting with maximum resolution and maximum frame rate. That figure would be a good benchmark for evaluating the efficiency of the IP camera and the calculation of the required network resources.
Apart from the encoder settings, this indication depends to great deal on the image content and the movement within the scene. With moving objects in the image, Progressive Scan offers the critical advantage of enabling outstanding image quality at low data rates. The independent tests published in magazines do not make any relevant statements regarding those criteria. Rather, standard test are being conducted using static images from the annals of video technology. For the reader it remains relatively unclear what methods have been used to test the encoders. Naturally, manufacturers of cameras and video recorders apply their own procedures, attempting to further expand their edge in quality.
The customer depends on the long-term reliability of the complete system and has a right to expect a corresponding quality and stability of the system. Although many discussions only revolve around the advantages and disadvantages of network cameras, one important point must not be forgotten. It is not just the individual components that are essential, but the reliable functionality of the complete system.
That is why many questions need to be answered in advance. What does the customer really need to solve his problem and to be able to monitor relevant areas around the clock? Are the recordings of critical scenes detailed enough and can they be used in courts of law? Does the system need to have specific certifications (UVV Kassen, Kalagate etc.)? Is the performance capability of the video management system sufficient to efficiently find video data and allocate as well as save accompanying information provided by third-party systems? Will the video system have to be integrated into an existing facility security system? Maybe there is an older system in place which shall be integrated but remain unchanged? What kind of investments will pay off for the customer, when, apart from the acquisition costs, any running charges and mid-term costs (wearing parts, maintenance) are included in the evaluation? Could it be recommendable to use specialised video analysis systems for automatic monitoring?
Only competent advice right from the beginning will ensure that the CCTV/IP system meets his requirements and be a secure investment in the long-term. The possibility of integrating older devices into a modern system offers the customer the flexibility of being able to use devices of different generations. The long-term availability of all spare parts ensures a high security of investment for the CCTV/IP system over longer periods of time. Also, the surveillance system grows with any building expansions and the ever-increasing requirements the customer puts on his individual solution.