Video analytics has matured enough to be a core piece of any CCTV installation. Much of the focus has been on the real-time detection capability of video analytics, but it also adds value in reducing storage and bandwidth requirements as well as reducing time needed for video review. However, in order for analytics to deliver maximum value, more training in the industry on best practices must be adopted.
Even as video analytics accuracy is still being improved, it is far more accurate than traditional methods of detecting activity such as PIR-motion sensors. Video analytics is a better choice over motion sensors in its ability to classify objects and events, even further reducing unwanted alerts. For example, in some cases it’s permitted to have vehicle movement but not people. In other cases the opposite is true. Another possible scenario could be left object or removed object detection. By using video analytics, operators have much more control over the specific types of events that cause alarms, giving them the ability to have a more focused approach to video monitoring.
In situations where operators are tasked in finding a specific event in hours of video, such as who damaged a car in a shopping mall parking lot, video analytics can be used to search video on the operator’s behalf and present them with a number of potential hits, much like Google is used to search information on the internet. The time savings are enormous, making “search” one of the biggest benefits of deploying video analytics. Studies have shown that harnessing video analytics in search can reduce the time required by 90%. Finding information faster not only saves money but makes the information found more actionable since the perpetrator may still be nearby.
With mega-pixel cameras becoming increasingly popular in CCTV deployments, the bandwidth and storage costs are quickly escalating out of control. However, video analytics can help by intelligently filtering unwanted video and only recording relevant events, much like anti-spam can filter junk mail from your inbox. This approach can keep storage costs in control while still allowing operators to have an acceptable number of recorded days of video.
Finally, video analytics can only be successful if the installer adequately understands the solution space and is properly trained. Lack of training and understanding is a concern in the industry. While the vendors need to offer better education and training, the installers need to show more interest in product training. Although video analytics can improve CCTV deployments immensely through reducing false alarms and better optimizing storage and bandwidth consumption, success is deeply rooted in the installer’s ability to properly configure the video analytics and choose projects within the intended solution space.
Justin Schorn is with Aimetis, a developer of video analytics technology.
