The role of video analytics (video analysis)

Please, note that this article is automatically translated from Russian into English, so the translation may not be accurate.

CCTV systems are used for surveillance, security and protection of strategic installations such as airports, railway stations, power stations. They are also used for commercial buildings (offices, retail, warehouse and manufacturing facilities), housing (apartments, cottages) and transport infrastructure.

The market of security systems, while keeping the popularity of non-intelligent surveillance systems that do not use modern advances in computer vision and artificial intelligence. These systems represent a certain number of cameras connected to the display system and recording device.

In the past decade has seen the development of surveillance systems and a sharp increase in the number of computerized systems.Videoanaliticheskie such systems can reduce the amount of participation in the process of observing and analyzing situations, the operator, which reduces the human factor - computers do not tend tired and distracted.

In addition, the computerized system handles a large number of incoming data in real time, rather than analyze it post factum.

The urgency of the problem

Many on the market today for surveillance use a tandem, which includes a video surveillance system as such and the operator. The need for operator intervention is dictated by the imperfection of the existing videoanaliticheskih systems, in particular - lack of process automation. The operator has to analyze the incoming data stream visually (manually), which is associated with very high physical and psychological stress. As a result of operator fatigue quality and, consequently, the productivity of observation and analysis are reduced, which negatively affects the efficiency of identification of abnormal situations.

If you use non-intelligent video surveillance systems with a large number of cameras installed, the chance to pay attention to emergency situations and at all tends to zero. This leads to the practice of analyzing data is not in real time, but in fact the event.

An urgent task is to carry out the structural upgrading of existing video surveillance systems.In particular this applies to automated processing and analysis of video streams. Solution could be the introduction and application of high-tech videoanaliticheskih CCTV-systems.

Video analytics (video analysis) as a way of solving problems

Called video analytics technology, computer analysis of video data in order to obtain systematic information about the object of observation without intervention. Typically, a video analytics software algorithms that can be run on computers and embedded in the camera.

Охрана стратегических объектов

Protection of strategic objects

Fighting crime and terrorism

Accidents prevention

The fight against impaired driving

Manual analysis of streaming video

Video analytics can efficiently solve problems such as detection of intrusion into the guarded area, counting, recognition of license plates of cars, the recognition of violations of traffic and many others. More complex tasks, which include such as face recognition, abandoned objects, detecting suspicious behavior, less successfully solved.

Introduction and use of video analytics systems in the areas of safety, security and surveillance is the most obvious solutions to security problems of a big city or the object of special purpose. Use of automated systems will reduce to a minimum participation in the process operator or completely eliminate the human factor.

Requirements for video analytics

High sensitivity detector

Analytics system should have a high accuracy of detection and classification of significant events for the subsequent reaction, regardless of environmental conditions (weather, lighting conditions). Algorithms for detecting and tracking should take account of volume to the scene and provide multiscale work in the foreground and the background.

High-speed detection

The analyst should be performed in real time (or within a short period of time) for immediate notification of the threat. This provides the necessary level of security.

To increase the speed of video processing is necessary to use processor and algorithms that can cope with the increasing workload and changing conditions.

High availability systems

Factor reliability is essential when using any security systems. Reliability can be achieved through design and engineering solutions. In addition, the system must be equipped with elements of passive safety, as well as secure data transmission through communication channels.

Low level of false positives

Members videoanaliticheskih systems often face the problem of false positives. Sometimes their number in the tens per hour, which leads to inappropriate use of such systems. Video analytics should be able to filter out the activity, no interest, which is important in the application of systems with many cameras (public transport, large buildings).

Resistance to changes in the environment

Videoanaliticheskaya system must possess the characteristics and resources that will use it in different conditions (weather, environment).

By changing conditions may include lighting (night, day, twilight, insufficient or excess lighting), weather conditions (fog, snow, rain) and so on. Regardless of
These factors aim videoanaliticheskoy system - to cope with the tasks.

Decentralization of processing, scalability

The need for embedded intelligence, which, unlike other solutions, you can create scalable systems whose dimensions are not limited to power and performance of the central node. System with embedded intelligence allows the transfer has already analyzed the data on any device video output (the computer monitor screen, PDA, etc.) or place of storage (remote disk).

Autonomy of nodes increases the reliability and security of the system - it does not depend on the health of a server.

Restrictions solutions

Widespread introduction of video analysis is constrained by the following factors.

The high rate of false positives

False positives common because of imperfect algorithms and methods for video analytics. They can be caused by the presence of rolling background (forest, water area), animals, birds, insects, found themselves on the camera's lens, as well as shake the camera itself and other factors.

Consequently, the system creates additional psychological burden on the operator and can be disabled as "annoying."

Complexity of integrating

Many products on the market analytics systems do not allow for their integration into existing platforms, or is it fraught with difficulties. As a rule, for the sake of integration is necessary to go to the additional financial costs.

The high cost of maintenance

Analytics systems are often a cumbersome solutions that require additional equipment for the task and the constant presence of a specialist. This leads to additional costs to the customer.

Often used by exclusive and expensive components of the software, which also affects the price.

Non-optimal placement without prior estimation of efficiency can affect the ultimate cost of the project towards its increase. The problem of accommodation as regards ensuring adequate level of security on the supervised area.

Name:
Email:
Phone: