Smart Vision using Machine learning for Blind

Abstract

Smart vision helps the visually impaired individuals to recognize and investigate the happenings encompassing them. According to the World Health Organization (WHO), about 15% of all the total population lives with some type of disability. In this total, visually challenged persons play a significant role. Their way of life is confined and they cannot look into themselves with the opportunities given to them. In any situation, individuals ought to totally rely upon their senses or others for discharging their day to day activities and this makes them lose their confidence. Until a decade ago, the aids that a blind person used were a stick, guide dogs or accompanying persons to move. With the advancement in technology a lot of support systems using Matlab, Radio Frequency Identification (RFID), Global Positioning System (GPS), and ultrasonic sensor modules were implemented. However these technologies resulted in few shortcomings and difficulties .To conquer these difficulties; this paper portrays a smart vision using deep learning. The proposed system comprises of Raspberry pi-3 processor, pi camera, tensor flow, memory card and audio amplifier module. This system definitely helps the visually challenged to identify the happenings surrounding them with at most ease.

Keywords: Smart vision, visually challenged, Tensor flow, Raspberry Pi, Pi Camera.