AI-driven video surveillance

Abstract

The threats to public safety are continually changing, which is putting a strain on law enforcement and requires new processes to support law enforcement. In this environment, footage of surveillance cameras can be of enormous value to public safety, but the human capital required to grasp all of that data is excessive. Recent technology based on Artificial Intelligence (AI) allows real-time surveillance footage to be analysed with fewer police officers involved. In Switzerland, such AI-supported surveillance cameras for real-time analysis are currently underdeveloped, despite success stories for public safety in other countries. Studies have analyzed the feasibility, public acceptance, and legal requirements for such smart surveillance systems, yet they have focused mainly on Anglo-Saxon countries. In this thesis, a body of research is analysed and extended with additional research to determine the potential, challenges, and risks of implementing smart camera surveillance systems (SSCS) in Switzerland.