Content-Oriented Surveillance System Based on ICN in Disaster Scenarios
Koki Okamoto, Toru Mochida, Daichi Nozaki, Zheng Wen, Xin Qi, Takuro Sato · 2018
This paper deals with an efficient image object detection method for use in a disaster prevention network that uses information-centric networking (ICN). In ICN for disaster prevention, a large number of surveillance cameras are arranged, and disaster image contents corresponding to the user's requests are distributed directly from the node attached to the surveillance camera. At this time, the name of the content requested by the user does not necessarily match the name of the image acquired by the surveillance camera. In this paper, the content requested by the user is processed and named using natural language processing (NLP). In addition, the image content from the surveillance camera is named using artificial intelligence technology. In this way, a method for improving the hit ratio between users and cameras was established. Furthermore, the volume of the interest packets decreases based on the information which area often occurs disaster. As a result, the network efficiency of ICN can be improved.