Towards a Framework for Collaborative Video Surveillance System Using Crowdsourcing
Susumu Saito, Tetsunori Kobayashi, Teppei Nakano · 2016
This paper proposes a new framework for video surveillance systems for crime prevention. The main purpose of this framework is to help provide reasonable and stable solutions for automated video surveillance systems in a collaborative way. This framework is characterized by a verification process using crowdsourcing after the image analysis process: automated image analyzer detects as many suspicious events as possible followed by filtering process using human intelligence, to achieve both high re-call and high precision rates. Here we describe the basic mechanisms for collaboration between camera devices, data stores, image analyzers and surveillance crowds.