Voice based Assistance for Visually Impaired Individuals

Abstract

This study provides a framework for a smart cane that offers an object recognition system for visually impaired people which assist them in a better walking experience. Framework will explain the details how Bluetooth enable device will identify the obstruction object and communicate it to the user through voice module.  For the working of the system, Ultra sonic sensors senses the obstructions clicked through moveable camera and transferred using Bluetooth   to the phone connected with the smart cane. Phone is installed with application that can process images using required API for ML (Machine Learning) and NLP (Natural Language Processing).  Outcome of the algorithm is a text that will be further converted into a voice alert and   communicated to the user through an earpiece.