Diseño e implementación de un sistema de navegación para asistencia de personas no videntes

Abstract

This work is focused on the implementation of a prototype that assists, complementary to the white cane, in the navigation of blind people by detecting obstacles with artificial vision. The navigation system was developed based on the Kinect sensor coupled to a helmet, which in addition to providing a color image of the environment, it is also capable of delivering a depth image of the same; this information was processed in a computer and based on the modified fictitious force algorithm an obstacle free route was determined, which is transmitted to the user through audible messages; furthermore a complementary vibration system was implemented and mounted on a vest, which will serve as backup in order to alert the person about near obstacles. In the same vest an interaction button was incorporated, which allows the user to have more control over the system as they can voluntarily activate or deactivate the audible messages emission. The prototype also has an identification and decoding system for QR codes that helps for a better orientation of the visually impaired person within the environment in which they manage themselves.