Video surveillance initially required constant observation by security or management personnel. Later, when systems had the ability to capture quality recording, much of the focus shifted to forensic review of incidents from the recording. Video monitoring systems, however, generate a huge amount of data.  Manual searches historically required a huge amount of time and effort to search through hours and hours of recorded video to find a particular event.  Not anymore.  Analysis algorithms can search either regions of interest or entire areas, and sift through mountains of data to quickly locate and pinpoint previously captured incidents.