Audio Augmented Reality Seamless Navigation Assistant Based on GNSS and Computer Vision for Blind and Visually Impaired People
摘要
The World Health Organization (WHO) reports that all over the world there exist around 2.2 billion people who are visually impaired or having low vision. Being this trend growing, the development of navigation assistants supporting Blind and Visually Impaired People (BVIPs) is of urgent need. Over the last few decades, researchers have developed systems that can make BVIPs more independent and aware of their surroundings. This contributes to create a human-centered built environment that prioritizes occupants, focusing on their needs, well-being, and experiences. Nevertheless, an integrated solution that can support BVIPs in both indoor and outdoor unprepared environments is still missing. To address this need, this study proposes an Audio Augmented Reality (AAR) navigation assistant for BVIPs that integrates Global Navigation Satellite Systems (GNSS) and Computer Vision (CV) technologies. Such a system, deployed as a smartphone-based application, enables high-accuracy user localization seamlessly in both indoor and outdoor scenarios and without any manual procedure. Furthermore, AAR technology, superimposing audio information into people’s physical 3D environments, is applied to provide audio spatial guidance to BVIPs. Experimental tests, carried out in a university campus, show the possibility for eye-folded users to follow a reference path with a meter-level divergence.