Smart Eye: Face Recognition & Obstacle Detection Aid for Visually Impaired

Abstract

Human life has advanced a lot using up growing technologies. There are a lot of difficulties faced by blinds in today’s society. The causes of blindness include diabetes, traumatic injuries, infections of the cornea or retina, glaucoma, and inability to obtain any glasses. In order to resolve this problem we are here with a glass with obstacle detection and face recognition. A smart glass is designed to help blind people to navigate and explore freely. The work starts by absorbing the scenarios around the person and detecting them using camera. The device thereafter recognizes the image captured and compares using the algorithms. The components required include processor Raspberry Pi, camera, earphones and a power source. The dynamic image is subdivided as frames in order to recognize the image. We make use of Tensor Flow API, open- source machine learning library developed by the Google Brain Team for the object detection and with providing data feed neural networks understands audio signals. Thus, the details of objects in a frame can be obtained by Tensor Flow API. A Text to Speech Synthesizer TTS software is used for converting the details of the object to speech format. Basic objects like cell phones, door, chair etc. will be detected. Camera-based face recognition has advanced to the point.