Often the most interesting footage comes into the control room late at night or from infrequently accessed areas and this algorithm allows the security team to know when people or cars are on site at times that they shouldn’t be.
Our usual time of day algorithm finds people and cars that are not expected in the scene at that time of day or night. Some examples of the types of events that this algorithm finds are:
Late night access activities
Cars & people in car parks at night
Maintenance staff on the premises when it’s closed
Access to restricted areas
The unusual time of day algorithm determines typical density expectations for people and vehicles for each hour of the day. A day is modelled from an earlier period that may be from the day before, a week earlier or a range of days of data. The system constantly updates this model to establish an expected number of people / vehicles in a scene at any given time and will raise an event when this number is exceeded.