IMAGO is a novel approach for guiding visually impaired people with the help of smartphones and image processing techniques. The aim of the project is to build a robust and affordable tool to provide positioning and navigation for blind and visually impaired pedestrians in outdoor environments. This method uses pre-recorded routes to compile a three-dimensional model linked with GPS coordinates and sensor data (e.g., GPS, accelerometer, magnetometer and gyroscope) on the route.
The navigational approach requires two steps: (i) route acquisition and (ii) navigation. At the route acquisition step, a person knowing the route has to walk it at least once. During this walk, the data required for the route model is acquired with the image recording app. The recorded information is embedded into the images transferred to a server. At the server, a 3D model representing the route is build from the images. At the navigation step, another smartphone application is used to pick the desired route to walk. After the walk initialization images and sensor data are acquired by the app and matched with a corresponding model. Identified locations of the matched images are used for the navigation.