Video management systems, armed with video analytics, can be a very effective means to detect and alarm on incidents. Despite the ability to accurately detect and notify, the issuance of an alarm does not always result in a timely resolution of the event. This may be due to the remoteness of a particular location, limitation in security personnel or delays from local law enforcement due to high activity. The automation of audio talk down, achieved by combining video analytics with audio, can help address this issue by providing an automatic, or semi-automatic, means of deterring an intruder.
The ability for a security operator to react to a security incident with direct audio is referred to as audio talk down. This feature enables the operator to directly inform the detected perpetrator they are being monitored. When an alarm occurs, the operator locates the target with a security camera and then issues live or recorded voice commands at the intruder. The use of audio talk down gives a virtual on site presence, even though the operator may be miles away. Often times an intruder may understand that cameras are present, but ignores the fact that real people are monitoring them. In other cases, they may recognize this, but expect they will have a great deal of time before any type of response may occur. In these cases, the use of directed audio has proven very effective.
Manual audio talk down has been used in security operation for many years. Communicating specific details, such as shirt color, location, current action has been shown to be very effective in deterring intruders. However, having a live person readily available to perform a talk down is not always feasible or affordable, especially for facilities that may have limited security resources or limited budgets for full time central monitoring. Even with full time monitoring, a person’s ability to react immediately may be delayed due to other circumstances or priorities. Combining audio with video analytics can help overcome these issues by partially, or fully, automating the audio talk down function.
A perimeter surveillance system that utilizes video analytics is an effective means of intrusion detection. An added benefit is that most video installations allow for alarming before the intrusion actually occurs (e.g. outside the fence). Video Analytics also has the ability to understand where the potential intruder is physically located on the property. Some solutions can classify the target as human, group of humans, vehicle and provide its current course or heading. Using that information, video analytics software can determine when to issue an audio message and choose from a set of prerecorded audio alerts, pseudo-customizing it to the event. Additionally, camera auto tracking, the ability for the video analytics software to control the pan, tilt and zoom of a camera to keep the intruder centered in the image, can be enabled to maintain visual coverage of the event. This can continue without any user intervention, or at any point, an operator may take control of the PTZ, the audio commands or both.
Most facilities are typically less concerned about apprehending the intruder and more concerned about avoiding damage, injury or theft. Understandably, intruders also want to avoid interaction. As such, the earlier they know that they are under surveillance, the more likely it is that they will leave, or at least make short work of their intrusion.
There are several methods to add intelligent audio talk down to an existing surveillance system. For a remote site or a site with limited cameras, a simple video analytics edge device can integrate with the existing video management system to intelligently monitor and control all the cameras at the location. For these installations, one or more speakers can be added, with the intelligent automated audio responses being sent from the video analytics edge device or user responses sent from a computer at the monitoring facility.
For larger or advanced installations, the addition of a video analytics server may be in order to accommodate more video / audio interfaces. For these installations, security may also consider the idea of a long-range audio deterrent (e.g. LRAD). An LRAD brings the added functionality of directional audio, whereby the audio in focused upon the intruder. When controlled via video analytics, the system can keep the directional audio pointed at the target, even when moving. As the name implies, these devices also have the ability to provide additional influence for the intruder to leave the area by asserting very loud, directional, deterrent tones.
Audio talk down has proven to be a very effective tool for protecting facilities and critical assets. Adding intelligence though video analytics effectively turns a passive surveillance system into a virtual remote first responder. This capability can be achieved with minimal additions to existing video management systems and creates a work force multiplier, which can aid remote facilities in reducing or eliminating damage, injury or theft.