Assistive Technology for Blind People for Object Detection and Description

Authors

Yasmin M A K, Reshma G, Saffar Subuhania T, UG Students, Edna Elizabeth N, Professor
Department of Electronics and Communication Engineering, Sri Sivasubramaniya Nadar College of Engineering, Chennai, India.

Abstract

The development of deep learning has helped object detection to make rapid progress. In recent years, smart wearable technology has become the part of everyday life. Smart wearable technology includes watches, glasses and many other wearable items. There are plenty of smart applications built for us. There is a lack of technology to aid visually impaired. Visually impaired people have to rely largely on other senses such as hearing, touch and smell in order to understand their surroundings. It is really difficult for them to walk without knowing what lies ahead even with a stick. Giving blind people the great accessibility to their environment is the objective of the smart glass system. In this proposed project, a smart glass application system for visually impaired people based on deep learning was proposed. This system can provide voice automated output which describes the obstacles along with the distance of the object from the user using an ultrasonic sensor which will help them to navigate without hindrance. The system predicts the class of the object and text to speech conversion is performed by using python library. It gives them an opportunity to visualize things and also guides them to move freely without getting injured.