A Video Analytics-Based Intelligent Indoor Positioning System Using Edge Computing For IoT

Abstract

In the context of Internet of Things (IoT) environments, obtaining the target's location information by analyzing a large number of video data captured by widely distributed camera nodes shows a good development prospect. As far as video analytics is concerned, for the purpose of reducing high processing costs and transmission time of video data, it is an effective solution to offload computing from the cloud to edge devices. In this paper, we design a four-layer video analytics architecture with the concept of edge computing and adopt the lightweight virtualization provided by the container technology to modularize the video analytics process. Based on the proposed architecture, an video analytics-based intelligent indoor positioning system is implemented. The proposed system can provide centimeter-level positioning accuracy and, at the same time, has a lower response delay than traditional cloud computing model. The experimental results show that, high-precision location information could be obtained with the help of billions of camera nodes, while for the large-scale video analytics, the usage of edge computing has great potentials.