When it comes to using video analytics for outdoor perimeter protection there are 2 types of technologies.
Standard video analytics has been around for more than 15 years and is the most widely used. This technology looks for the change in pixels to determine if there is activity. It differs from standard VMD or video motion detection because it typically looks for the pattern of pixel changes to determine if the motion is caused by people, cars, or something else. If a high enough percentage of pixels change in a defined pattern the system triggers an alert. The advantage of this tech is the speed it can detect an intrusion. The disadvantage is that it creates a lot of false alarms.
The newer technology that has evolved and is currently evolving is Deep Learning video analytics. With Deep learning the system needs to have a clear view of the object (close up). Deep learning can identify the motion as male/female, approximate age, assess threats and many other advanced functionality. At GoView we call this Video AI. The best systems use both technologies to quickly identify activity and self verify the motion to reduce false alarms. All systems will produce false alarms but the rate of false positives can be dramatically reduced with the right technology. Contact GoView to learn about how our Video AI solutions can help you solve security problems