Organization of the workplace of the video surveillance operator with a large number of IP cameras.
Video surveillance has been the most informative, most visible security subsystem for several decades. Yes, alarm systems have the highest specific efficiency per ruble of investment, but once the alarm is installed, no amount of expansion will greatly increase the security of the facility. Video surveillance allows you to process a huge amount of information, and, as a rule, the more video cameras are installed at the facility, the higher its security. However, a large number of video cameras, a large amount of information is a sea in which you can drown if you do not organize an effective system for managing this information.
The simplest way to help the operator is to select video cameras for display not by their number, but by indicating them on the site plan — this significantly simplifies management in a critical situation: you do not need to frantically remember the camera number or look for it in a long list hanging on the wall near the operator.
In addition to selecting single cameras, it is often useful to organize the so-called salvo output of several signals from one section (from one room) to several monitors. On a computerized plan of the room, the operator selects a room with the mouse (or a finger on a sensitive screen), and several cameras from this room are displayed on the video monitors at once. The distribution of cameras on these monitors must be meaningful and uniform. This is often easy to implement, for example, images from several video cameras located in a row on the perimeter are displayed on several monitors installed in a row. In this case, the correspondence of the cameras to the monitors is obvious. Sometimes the correspondence is not so obvious, but if several (but not more than 4) video cameras are installed in all the rooms, then the operators quickly get used to the fact that this monitor shows a view of the entrance door on the left, this one always shows a view of the window on the right, and this one shows a view of the entrance door from the outside, from the corridor. But for this to work, the cameras must be installed uniformly in all the rooms. Even if some rooms do not have the full usual set of video cameras, nothing should be displayed on the monitors corresponding to the missing cameras, and the existing cameras should be displayed on the same monitors to which the operator is accustomed. In extreme cases, a technique that has been seen may come in handy: when switching several cameras at once, they are displayed on the monitors deliberately slowly, with short pauses, and as they are displayed on the screens, they are highlighted on the room plan. In this way, the operator receives information about which camera is displayed where. However, I do not think that he will be able to remember this for more than a few seconds. I repeat once again — the only chance to effectively use the display of several images is to lay them out on several monitors (windows) uniformly, in an obvious way if possible.
In fact, pure video surveillance with a large number of video cameras is rare. Since the facility is so complex or important that it has many video cameras, there are probably other security subsystems — alarms, access control — so that the operator's workplace is essentially integrated, allowing you to control the integrated security system. Even if the dominant system is the video surveillance system, the system will still be integrated to some extent. By the way, the computer video surveillance systems that exist today, or systems supported by computer workstations, are always integrated. Previously, there were hardware systems on matrix switches for tens of thousands of video cameras (for example, in some airports), but now computers with integrating software are used everywhere to control the systems, and even old analog video systems are increasingly integrated with security alarms, at least through built-in video switches and video multiplexers.And this is significant, because it is unrealistic to manually monitor hundreds or thousands of video signals — this would require tens or hundreds of operators. The system itself must select the most interesting (dangerous, suspicious) images and show them to the operator.
Thus, video surveillance is not an independent system, it is designed to verify alarms or control the development of a situation when other sources of information indicate that there is a problem in a certain area and the operator manually views only a few video cameras — one area.
However, based only on the data from security alarms and other systems, it is quite difficult to choose which surveillance channels are of most interest at the moment. Automated analysis of the video signals themselves can help, but in any case, the system will only be able to highlight several video cameras as potentially the most interesting, and the operator must manually determine which channel (or several channels) deserves careful study. For example, a glass breakage signal may be of interest to several video cameras inside the building (if the offender has broken in) or several outside (if he is running away). Even a signal from a motion detector in one of the cameras often does not provide information about which camera’s field of view the intruder will appear next, because he has already run through that camera’s field of view, and there is no point in looking there – except to review the recording for the last few seconds. Thus, the main postulate of organizing an operator’s workstation with a large number of video cameras is the ability to quickly glance at several images, select one among them, study it closely, switch to the next one, and back. In a primitive case, this is implemented in the form of a multi-window display of several signals, from which the operator selects one for careful consideration on the entire screen. In a primitive way, because with such a workplace organization, after selecting one image for viewing, the others are no longer visible to the operator and he may not notice that more interesting events have begun there than in the channel that he is currently viewing in full-screen mode. If the designers are not very limited in the dimensions of the operator's workstation, it is better to use several small monitors (or one large one in multi-window mode) as overview monitors and one high-quality one for careful study of the selected channel. Some systems allow finger selection on the sensitive screen of a multi-window overview monitor. I have happened to design systems from several simple monitors with buttons installed next to them, configured to quickly display the image from this auxiliary (overview) monitor on the main monitor.
Psychophysiology allows us to scientifically justify the choice of the configuration «one large monitor and several small ones». A person has a focus of attention. There is little point in placing several identical monitors for careful examination, since the operator will only be able to look closely at one anyway. However, the operator is able to see the other monitors out of the corner of his eye or periodically quickly run his eyes over them. These monitors are not required to be of the same high quality as the main one, but they should be located compactly so that they fall into the field of view out of the corner of the eye or at least can be glanced at without turning the head.
For reference: the angle of active human vision is approximately 15 degrees in all directions from the axis of vision. The angle of vision out of the corner of the eye is 30 degrees to the left, right, up, and down from the axis. The angle of vision without turning the head (by turning the eyes) is 60 degrees to the left and right, 40 degrees up, 50 degrees down. By the way, the gaze in a calm state is directed 15 degrees below the horizon. Thus, it is optimal to place all monitors in the area of 30 degrees left and right, 15 degrees up from the horizontal at eye level, and 45 degrees down. In this case, the central monitor should fit into the central ± 15 degrees. In extreme cases, it is permissible to use an area of up to 60 degrees left and right, 25 degrees up, and 65 degrees down for unimportant data that will not be perceived out of the corner of the eye.
The main monitor for careful viewing should be as convenient and high-quality as possible. Since the most common video format is still 625 lines, one picture on modern HDTV monitors may not occupy the entire screen. It is permissible, for example, to divide a large monitor (with a resolution of 2000 x 1500) so as to select an area with full quality of 800 x 600 and several smaller windows, and also to display related information from other systems on the same monitor. The advantage of a single-monitor solution is the ability to dynamically redistribute the screen for different tasks. The disadvantage is that the inconstant division of the screen makes the operator's work difficult, and in addition, if this monitor fails, the workplace becomes completely inoperable.
The image quality on the main monitor should be limited only by the video camera. Therefore, the lines should be strictly on the edge of the eye resolution — about 1 arc minute. With a higher value, the lines will be visible and will irritate the operator, with a lower value — some information will be lost, the resolution will be limited not by technical means, but by the operator's vision. Thus, for example, an analog video monitor with a diagonal of 17 «(43 cm) should be located at a distance of about 140 cm. In reality, the distance may be less, because the vision of most people, especially after several hours of duty, is far from ideal, in addition, high-quality black and white analog monitors do not have clearly defined lines. Although in low-quality monitors, the lines of even and odd half-frames can be doubled, so that the actual size of the line will be twice as large as the theoretical one. If an LCD monitor or a high-quality high-resolution computer monitor is used, then it is necessary to apply computational smoothing algorithms and output one line of the video signal to 1.5–2 physical lines of the monitor, then the image will also become smooth and the screen can be brought closer to the operator, which is more comfortable, without the risk of eye irritation from the visible line structure.
It is especially important to hide the discrete structure when displaying computer text and graphic information. The monitor resolution should be one and a half to two times higher than the eye resolution, so that the eye does not see the denticles on the lines and dots on the letters. Therefore, if text (graphic) and video information are displayed on one screen, then at least 1200 x 900 pixels will have to be allocated for the «main window», and even more — 1600 x 1200, which completely eliminates any savings from attempts to combine the controls of the computer program and the video image on one screen. By the way, since we are talking about image quality, I will note that many people still design analog video monitors to display the video signal for another reason: this allows you to minimize the number of analog/digital/analog conversions — an analog monitor with an analog camera reproduces the signal most accurately.
Another consideration is not related to the organization of a separate workstation, but to the organization of the entire system. If there are hundreds (or even thousands) of video cameras, one operator will obviously not be able to cope. Conclusion: it is necessary to build a hierarchical system in which the main work is performed by local posts (controlling nearby 10-100 cameras and corresponding elements of other subsystems), and central posts for centralized support of mobile groups and comparison of the situation in different areas.
A special type of multi-channel observation posts are situation centers. These are rooms designed for many different specialists, who are in a critical situation. For example, the Moscow Main Directorate of Internal Affairs has a room where dozens of people can work simultaneously, coordinating the actions of many departments. Video signals from thousands of video cameras can be output there. Video information is the main tool in crisis management centers. Video information often carries much more information than any sensors — the degree of panic can be understood by the expression on people's faces, different specialists can isolate important information for them about the state of technological structures, the degree of fire development, the adequacy of the availability of medical workers, etc. In the end, crisis situations are crisis situations because they are non-standard, it is not known in advance what information may be needed. And video information is the most universal, it is not for nothing that a person receives up to 90% of all information about the world around him through vision.The situation center usually implies a controlled (edited) display, almost like on television. There is an operator who, like a television editor, decides what to show on the main screen (or several screens), and what to put on auxiliary screens. Many specialists present in the room have their own workstations, their own monitors, on which they can examine in more detail the picture that interests them. But the essence is the same: one or two large screens show the main moments of what is happening, and several smaller screens show what is interesting only to certain people present or what may unexpectedly arouse interest in a certain development of events.
A similar case is technological systems. In a normal situation, the duty operator (the duty shift of operators) monitors the rolling mill, hydroelectric power station or cracking columns, etc. Most of the console is occupied by special technological mnemonic diagrams displaying the state of the installation. But a noticeable place is also occupied by video monitors for displaying everything that even slightly goes beyond the normal technological process. Everything that was not provided for can be observed not by means of special measuring sensors, but only visually, on the video monitor screen. And like the situation center in the control room (dispatch) of a large plant, if necessary, all the management gathers, and therefore the monitors are designed not only for observation by one full-time operator (as in the security service), but also for the presence of observers «in the second row». However, the basic principle remains the same — one main monitor is needed (it can be parallel with the main operator monitor) and several overview monitors (these can be the same ones that are overview monitors for the operator in normal operation).