Informações Principais
     Resumo
     Abstract
     Introdução
     Conclusão
     Download
  
  
  
 
Abstract
 
 
Acadêmico(a): Jean Carlos Franco
Título: Aplicativo para reconhecimento de objetos em imagens destinado a pessoas portadoras de deficiência visual
 
Abstract:
In Brazil, there are approximately 35 million people with some level of visual impairment. Of this amount, more than 6 million people have great difficulty to see and about 500 thousand people can't see anything, a condition that can generate difficulties in carrying out daily tasks. Because of this fact, this work presents the development of a mobile application in the form of assistive technology for the recognition of objects in images, being intended for people with some level of visual impairment. This application is intended for the Android platform and was developed using Java programming language. It also uses Google Cloud Vision API to perform analysis of an image and present to the user details of its content. Since it is intended for the visually impaired, the application presents the user with voice instructions at each step performed and describes the content of the image through speech synthesis. The application was able to convincingly describe about 67% of the objects and scenes analyzed, demonstrating the potential of the proposal of an assistive technology and the assertiveness of the Cloud Vision API. However, the tests were performed on users blindfolded, not on the visually impaired. The results achieved were satisfactory and the objectives of the work were achieved.