| Popis: |
Indoor navigation has many practical applications. But since GPS device outputs are not available inside buildings, a new technology is needed. In our approach indoor navigation is image processing based, in a multi-user environment: a multitude of smartphones interacts with cloud servers. This paper refers to one of the biggest challenges of such an approach, viz. to efficiently implement a scalable infra-structure for large volumes of image information in real-time. Results are presented and validated by simulation of multi-users and actual measurements for a variety of situations of a single user. |
