An Assistive Indoor Navigation System for the Visually Impaired in Multi-Floor Environments

This paper presents an innovative wearable system to assist visually impaired people navigate indoors in real time. Our proposed system incorporates state-of-the-art handheld devices from Google's Project Tango and integrates path planner and obstacle avoidance submodules, as well as human-computer interaction techniques, to provide assistance to the user. Our system preprocesses a priori knowledge of the environment extracted from CAD flies and spatial information from Google's Area Description Files. The system can then reallocate resources to navigation and human-computer interaction tasks during execution. The system is capable of exploring complex environments spanning multiple floors and has been tested and demonstrated in a variety of indoor environments.