Selecting the architecture and structure of a comprehensive automated recognition system.
Selecting the architecture and structure of a comprehensive automated recognition system.
Biometric features are clear, individual, biologically determined characteristics of each person. In principle, no two people have the same biometric features. Examples of biometric identification methods include fingerprint analysis, geometric shape of hands, iris or retina pattern, location of blood vessels, thermal image analysis, face, voice, signature dynamics, keyboard rhythm, etc.
Face recognition is perhaps the only biometric method of identifying persons that does not require special equipment.
The face recognition method is the only biometric method for identifying people and in terms of multi-purpose application. Unlike other biometric methods, which are only applicable for access control or database comparison, image recognition technology allows detecting a person's face in a video frame for subsequent image comparison with a database. Thanks to built-in infrared emitters, the computer easily recognizes a rubber mask as a dummy simulating a face. The observed object must have human skin, natural facial expressions and be «alive», otherwise an audible warning signal is triggered.
Recognition by facial features occurs at a distance, unnoticed, without attracting a person's attention. From the point of view of security services and special services, this is an undoubted advantage. In addition to security services (enterprises, airports, supermarkets, casinos, banks), the main consumers of such biometric systems include government agencies (ministries, law enforcement agencies, special structures).
The stage that must be completed when building a complex automated recognition station (CARS) is the choice of architecture and structure of the developed station. To solve this problem, it is necessary to consider the composition of the equipment, technical characteristics and mechanisms of interaction of the system's components.
The following subsystems must be included in the CARS:
— video surveillance to ensure continuous (24-hour) monitoring of the situation in buildings and structures;
— facial recognition systems.
The tasks can only be solved using digital technologies. Priority in the implementation of digital technologies should be given to the video surveillance system, as the most effective security system. This will implement the following functionality:
— transmission of live video over both local and global (Internet) networks;
— playback and viewing of video recordings from remote workstations;
— preservation of video recording quality when reusing the storage medium;
— instant search by time, date, data from motion detectors.
— maintenance of multiple channels;
— control of video recording quality;
— ability to add new functions;
— modular design;
— communication via high-speed port;
— control of cameras from the menu on the monitor screen;
— control of operating modes of systems, viewing video cameras and archives from both local and remote workstations;
— access to the system by password (switching on, switching off, recording, viewing);
— non-stop circular video recording;
— connection of any access control systems, fire alarm systems, notification systems that have open protocols for interaction with other systems, as well as connection of systems registered in existing system modules.
With the transition to digital technologies, it becomes possible to implement additional systems:
— face control, with the ability to professionally recognize a face;
— a promising direction for identifying threats based on human behavioral characteristics using three-dimensional detection techniques;
— intelligent search engines;
— detection of abandoned objects, etc.
Organization of control of technical subsystems of security provision from a single centralized control point.
One of the most significant advantages of an integrated security system is the possibility of organizing centralized control of the entire complex of subsystems from a single centralized control point.
Depending on the need to organize access levels and allocate management functions of subsystems, each user and specific automated workstation is assigned an access level and allocated a certain range of possible management functions.
In addition to vertical level division when organizing automated workstations, it is possible to create workstations based on the unification of systems by functional purpose. For example, a workstation is organized from which it is possible to control the system as a whole by objects, similarly, centralized control is organized along the line of the video surveillance system, etc.
In addition to the vertical gradation of management capabilities from a specific automated workstation, it is also possible to organize horizontal interaction between workstations.
Expansion and change in the configuration of the structural diagram of the organization of management of an integrated system with the organization of multi-level automated workstations is carried out in the process of adapting the system to changes in the requirements for ensuring comprehensive security
Advantages in ensuring comprehensive security in the case of creating an integrated system.
The obvious advantage is as follows: — the ability to centralize control of a complex of security systems from a single workstation; — creation of an archive of all events in the system and a long-term storage video archive; — instant retrieval of video information from the archive by date, time and event; — implementation of face recognition technology, face localization in the stream; — organization of interconnected work of all technical security subsystems based on the implementation of the principle of support and enhancement of the resulting capabilities of interconnected subsystems; — the ability to organize automated control workplaces of various levels; — the ability to horizontally and vertically combine controlled objects; — receiving video information and information on the status of the security alarm from remote objects; — the ability to organize video surveillance on a selective principle, as a consequence — the rejection of “dumb” video surveillance. Submission of information and video recording according to previously programmed commands; — the ability to modularly expand the system depending on the need for updating and implementing additional functions; — strengthening the security of protected objects; — organization of total control over the situation at the facilities.
System solution based on the network.
Video recorders transmit their images to a central server. All images are registered in one large video database. Additionally, several local viewing workstations can access the same server in parallel with the registration of images. A RAS server (remote access service server) offers users access via communication networks (e.g. ISDN).
The main disadvantage of this architecture is that in the same network segment it is necessary to transmit frequency bands containing data for the recording device, while other user information contains, for example, data necessary for playback (viewing) of the image.
Architectures of a large and complex video surveillance system with several video recorders (recording clients), several local viewing/installation stations (viewing/installation clients) and several video servers
Network segmentation consists in separating network segments with a high video information load from segments with other user data. Video information viewing represents a much lower network data load, since in this mode data transmission is less regular and slower than in the recording mode. The routing capability of the TCP/IP protocol solves data transmission issues within a segment without problems. Moreover, it allows changing the network type (Token Ring, Ethernet, ISDN) from segment to segment. Routers or «gateways» forward data to the next segment and, if necessary, adapt the communication protocol between networks of different network types.