An indoor navigation app using computer vision and sign recognition

Abstract

Indoor navigation is a major challenge for people with visual impairments, who often lack access to visual cues such as informational signs, landmarks and structural features that people with normal vision rely on for wayfinding. Building on our recent work on a computer vision-based localization approach that runs in real time on a smartphone, we describe an accessible wayfinding iOS app we have created that provides turn-by-turn directions to a desired destination. The localization approach combines dead reckoning obtained using visual-inertial odometry (VIO) with information about the user’s location in the environment from informational sign detections and map constraints. We explain how we estimate the user’s distance from Exit signs appearing in the image, describe new improvements in the sign detection and range estimation algorithms, and outline our algorithm for determining appropriate turn-by-turn directions.