Traditional video monitoring using analogue CCTV cameras suffered from limited switching and viewing capabilities. It was also limited by its low storage and poor retrieval capabilities.
The move to digital monitoring and recording systems overcame some of these limitations but proved to be inadequate for handling installations with a large number of cameras. The biggest obstacle was the requirement that security personnel are fully alert and completely attentive to their video screens at all times.
According to Patrick Meaney, founder of Video Analytics. Info, “Research shows that after about 20 minutes the average CCTV control room operator takes in about 5% of the information presented”. The likelihood of spotting a real event is estimated to be around 1 in a 1000. The review of recorded information is equally impacted by fatigue and information overload. As with many other walks of life the “Human Factor” plays a crucial role in deterring and combating crime. “Video Analytics is the most important breakthrough in the application of CCTV”.
CCTV systems have had Video Motion Detection (VMD) capabilities for many years. However, these have had very limited operational functionality and have been rather prone to generating false alarms. Among the factors that contributed to the high false alarm rates were – changes in lighting and weather conditions, both during a day as well as over seasons. In addition, these systems were totally incapable of handling complex scenes.
How does Video Analytics differ from Motion Detection? It is more discriminatory where false-positive triggers is concerned. – The user can specify the activity of interest not just movement.
Event Detection is based on a growing range of criteria: Initially – Non Motion, Direction, Size of Target, Position, Dwell Time, Speed of Travel, Distance Moved, and Colour.
With Motion Detection you can have several zones within an image. With Video Analytics you can set an area threshold for the target object inside the target zone. Scene depth can be considered for the first time with high quality real time digital video.
There is growing use of Video Analytics in virtual tripwire/sterile zone monitoring, people/vehicle counting, tailgating in Access Control.
To configure a Video Analytics system to cope with fluctuating environmental factors such as lighting, weather, change of seasons you need a Video Analytics system that has adaptive algorithms to ensure system accuracy and reliability.
It is important to understanding and manage risk factors in implementation:
- Does the software do what you need it to?
- Is it configured correctly for the application and environment?
- Have the operators been trained to use and configure the system?
Do you need Video Analytics as part of a CCTV security solution?
- If you want event detection to be automated
- Early response to threats to prevent escalation
- More effective and efficient use of resources
You do if you want to do
- motion detection & object segmentation
- object tracking/people counting
- object classification
- image masking
How should cameras be deployed for an effective VA system?
- This will depend vey much on the application and environment
- A radius arc of 60m is advised for “virtual tripwire” type doorway or perimeter sterile zone alarm
- Greater distances are possible but affect evidential quality of the recorded images.
Video Analysis results in cost benefits of reduced monitoring time & faster response.
Now, we can talk about security in terms of “return on investment” not “imposed purchase”. Video intelligence can bring many benefits and cost savings to an organisation
With the increase in computing speed and power and the development of sophisticated algorithms to analyse the images from the CCTV cameras Video Analytics became a reality. While it did overcome the limitations of traditional VMD systems, in the initial stages, the Video Analytic capability was overhyped and oversold. The vendors failed to deliver on their exaggerated claims and the buyers found that they had been taken for a ride. The outcome of this has been that in a number of video analytics enabled systems the capability has been switched off due to the constant high false alarm rate.
Even today the technology is being overhyped and oversold. The industry needs to educate the customer to have more realistic expectations. The use of Video Analytics is growing rapidly. It is important to understand its capabilities and benefits.