Desenvolvimento de protótipo de sistema para auxiliar a locomoção de deficientes visuais no ambiente urbano utilizando sensores e visão computacional através do RASPBERRY PI.

Abstract

This work aimed to develop a prototype capable of assisting the locomotion of visually impaired people during a walk in an urban environment. Initially, the theoretical basis for the development of this work will be presented: concept of visual impairment, accessibility and mobility, research on the difficulties faced by people with this disability even using a walking stick or guide dog, study on the ultrasonic sensor that detects obstacles in adverse environments. In addition, the Python programming language is explored together with the OpenCV library for the use of Computer Vision technology, aiming to use the object detection method, YOLO, and also to calculate the distance between the visually impaired and the detected objects. Subsequently, the steps and materials necessary for the construction of the prototype are presented, followed by the tools used, a detailed description of the experiments, and implementation of what was mentioned above. As a result of these experiments, images are displayed showing the detected objects and the distance they are from the visually impaired. Finally, the data obtained showed that the prototype developed meets the proposed requirements by informing the visually impaired about the obstacles around them and how far they are from them, significantly helping them during a walk in an urban environment.