The great problem with these cameras is they are not having the deterrent effect that law enforcement and private concerns had hoped for. According to author Patrick Radden Keefe, in his August 22, 2016 New Yorker Magazine article “Total Recall” , “CCTV was clearly effective at recording malfeasance, but … if so many people were committing crimes on camera then perhaps the cameras weren’t much of a deterrent… many rational criminals had clearly concluded that the police were unable to do much with the CCTV footage.”
This represents one of the best applications of deep learning and vision processing at the camera where intelligence can quickly identify an illegal act, medical emergency, accident, natural disaster, or other out of the ordinary event and immediately alert the appropriate first responders. In the case of criminal activity, the intelligence can access a database in the cloud of mug shots or passport, green card, driver license… photos and alert the police to the likely perpetrator(s). This ThinCI deep learning and vision processing functionality can easily be added to any networked camera at a modest cost per camera.