Intelligent video surveillance

From Wikipedia, the free encyclopedia

The creator of this article, or someone who has substantially contributed to it, may have a conflict of interest regarding its subject matter.
It may require cleanup to comply with Wikipedia's content policies, particularly neutral point of view.
Please discuss further on the talk page.

This page has few or no links to other articles. (Tagged since May 2008).
You can improve this article by adding links to related material, within the existing text. For some link suggestions, you can try Can We Link It tool. (You can help!)

1 Introduction
2 Evolution
3 Fundamentals
4 Applications
5 Advantages
6 See also
7 References

[edit] Introduction

A video surveillance system consists of networks of closed-circuit television (CCTV), multiple display monitors and a team of security personnel. The deployment of CCTVs has to be carefully considered to provide a full coverage of key areas to be under surveillance.

An Intelligent Video Surveillance (IVS) System involves the use of video analytics in video surveillance system which automates the decision-making process of humans in identifying potential threats. The word “intelligent” signifies the use of smart programming software as assistance to humans during the surveillance process.

[edit] Evolution

The first generation video surveillance involves CCTV cameras using video tapes for recording. A multiplexer serves the purpose of collecting various images for displaying on monitors which are monitored by security officials only.

The second generation video surveillance replaces video tapes with digital video recorder (DVR). Automatic event detection enables recordings to be triggered by motion events in the scene to save up memory space.

The third generation video surveillance involves more advanced technologies and the integration of intelligent algorithms. IP cameras were used in combination with network video recorders. Algorithms were programmed to perform object detection, tracking and event analysis which greatly enhanced the applications of video surveillance.

[edit] Fundamentals

IVS System under operation.

A video consists of numerous frames of images that integrate together to form motion pictures. Such images involve a finite number of pixels that contribute to the formation of the entire scene.

One of the most common events that occur in a scene under surveillance is motion. The idea of motion detection is to break such scenes into pixels in order to perform an analysis. Video analytics consists of logical calculations of such pixels so that if they changed in a directional manner, a motion of a particular object is detected as a result.

[edit] Applications

There are numerous analytical functions available based on different program codes. Various events include the detection of individual behaviours (Intrusion, tail-gating, suspicious motion etc.),crowd management (Line control, Counter-flow etc.) and object status (Static object, stopped vehicles etc.). Detection algorithms are highly customizable to different IVS Solutions depending on the requirements for particular projects.

Perimeter Security is embedded into most IVS solutions. The concept is to detect moving objects that enter a pre-assigned detection window, and to alert the controller upon such events. The algorithm can be applied to security zones, restricted areas, private areas etc. The usual implementation targets may range from jails, parking lots, private houses or even war-fields.

Apart from security, IVS systems can also be used for statistics. Behaviour Analysis monitor human's reaction under certain environment or situation. Retails shops install IVS solutions to monitor customers' behaviour such as walking pattern and duration of stay in front of particular shelves within the shops. Such statistics are extremely important to stores designs and inventories allocation.

[edit] Advantages

The ability to analyse and customise parameters in IVS Solutions.

- Real- Time Video analysis can be performed in real-time, enabling the IVS system to issue warnings automatically once potential threatening activities are detected.

- Functionality IVS algorithms are highly customisable. Systems can be updated with different event detection codes, providing a wide range of functions to the existing infrastructure.

- Automation The traditional method of video surveillance required security officers to monitor several scenes. Due to the high rate of human error, this puts the surveillance in jeopardy. Using intelligent algorithms can accurately identify pre-programmed human activities, which issued alerts upon such events.

- Accuracy The existence of humans in a particular system increases the error rate. The installation of IVS systems as a replacement to manual surveillance reduces error rate by humans who often fail to identify threats from the numerous scenes under surveillance.