 |
| Security professionals and end users are eager to define video analytics |
In the era where “I” is king, video analytics have become a topic of excitement, confusion, and deliberation. All one needs to do is take a stroll through the latest security expo to recognise how popular the space is, and how difficult it is to find an adequate explanation of what they are and how they work. So what are video analytics? This is the question both end users and security professionals are eager to define, yet depending on who they talk to, they may get different answers.
But the bigger question being asked is how can video analytics improve my security? Cale Dowell, Regional Director of Business Development for THRIVE Intelligence™, explains that the key issue here is not just a definition, but also an explanation of where video analytics can make an impact and how they can be applied.
Video analytics – Sorting through the ambiguity
Interesting to note is that video analytics are not as new as they seem. They’ve been in the marketplace for several years, primarily overseas, and are more commonly known as Video Content Analysis (VCA). You might be surprised to know that your child has an application of video content analysis in his or her Xbox Kinect. Other industries have used components of video analytics to help with supply chain management. But in the security industry, there seems to be some confusion as to how they work.
You may have heard definitions such as motion detection, pixilation change, intelligent sensors, processors, or even (my personal favourite) “magical” cameras. The reality? It has nothing to do with magic.
Real video analytics focus on analysis and classification of a live video stream. It is not a simple motion detection camera. Good analytics will utilise more than pixilation change as part of the classification process. Generally speaking, real video analytics operate in three general ways.
First, we start with the camera. Any security camera is designed to act like a human eye: It captures images in real time as a data stream. The first thing real video content analysis accomplishes is learning the field of view, or the normal data that is transmitted. We’ll call this the background. Once the background is established and separated (background modelling), any new information (objects) in the field of view are identified immediately. At its most basic core this is how video content analysis works. It learns the background data so it can distinguish new information that may interrupt the learned space.
Second, video analytics analyse the new information and compare it to pre-classified objects based on a variety of metrics depending on the sophistication of the analytics. Good analytics will utilise metrics such as size, shape, speed and trajectory within the field of view of the camera, in addition to pixilation patterns. The data is analysed by processing the information in much the same manner as a computer processes information.
 |
| Real-time perimeter detection is one of the key uses of video analytics |
Lastly, good video analytics will provide the ability to create zones, lines and triggers within the field of view of the camera. This allows for a custom-tailored approach to creating protocols for when/why/how to be notified.
To summarise, true video analytics are not simple motion detection cameras. They are more intelligent and operate in a manner that also analyses information as opposed to simply detecting motion or pixel change. This permits a higher level of customization and a more significant reduction of false alarms than previous technologies. The value is largely focused on receiving real-time alerts when the analytics are triggered, and eliminating as many false alarms and false positives as possible through proper setup.
Value proposition
The true value of video analytics rests primarily in their ability to significantly reduce false alarms and increase real time response. Where security cameras have been used for decades as reactive tools, this new technology is shifting the industry towards real time intervention.
By pairing real-time analytics with boots on the ground personnel, an end user gains maximum efficiency in monitoring and responding to events as they occur in addition to increasing the capabilities of personnel and site intelligence. The key fundamental uses of video analytics are:
- Real-time perimeter detection
- Force multiplication of on-site guard force
- Augmenting manned security labour
- Creating metadata based on image sampling
End users are beginning to compare the value of spending high dollars for manned live monitoring versus the ability to leverage video analytics to eliminate or repurpose the stationed monitoring guard altogether. Increasing coverage and monitoring is much more scalable and cost effective than simply using manpower alone. Many end users are realising significant cost savings by applying hybrid solutions (manned security in conjunction with real time monitoring using analytics) across their regional and national footprints.
Understand the ROI
Make vendors work for you by requiring a customised, consultative approach that delivers their full solution based on your needs, vulnerabilities, threats, and daily operations early in the vetting process. Disregard vendors that operate in a “one size fits all” mentality |
As with any investment, end users shouldn’t let the excitement overrule due diligence. With video analytics on the rise in a rapid way, they should be wary of companies offering substandard solutions for top dollar. Thus, it is imperative to understand the capabilities and solutions being offered in addition to drilling down to the bottom line of how it might save you money. And the reality is, in some cases it won’t save you money. In these scenarios, sometimes the cost will be well worth it, and sometimes it won’t.
As an end user, here are some practical questions to ask and consider through the vetting process:
- What are the costs and ease of use?
- What can your analytics accomplish for the security of my facility?
- Can I see a live demonstration of your analytics?
- How do your analytics work?
- Note: if the answer is simply motion detection or pixilation change, they are not true analytics.
- Where is your monitoring centre and can I take a tour?
- Note: any good monitoring company should always give you a tour if you ask.
- What is being outsourced and what do you manufacturer and design yourselves?
The Complete Package
The real value of analytics is applied through real-time monitoring. Big players in the industry are edging to prove their value by demonstrating end-to-end solutions pairing remote video monitoring utilising analytics with other traditional services such as access control, intrusion alarm monitoring, and conventional guard services. This hybrid solution is the future of physical security. This type of scenario can maximise capabilities, increase response time, and improve cost control. Make vendors work for you by requiring a customised, consultative approach that delivers their full solution based on your needs, vulnerabilities, threats, and daily operations early in the vetting process. Disregard vendors that operate in a “one size fits all” mentality.
Remote video monitoring that deploys true analytics is the catalyst for overwhelming change within the industry. But the true value is found by leveraging a meeting of the minds that is flexible to adapt, change, and grow with technological advancements.
Remote video monitoring can seem like a frightening undertaking, especially for large corporate users. However, there is significant value to be gained by offloading, augmenting, and supplementing your security needs through qualified third parties that are specifically tailored to meet your needs. Most CEOs are determined that they did not start their business to also run a security company. As a result, big box businesses are looking for viable solutions that can improve capabilities and reduce cost now more than ever. True analytics are simply the current linchpin that is allowing that desire to become a real actuality.
