EBSAR: Detecting of Objects that Hinder Visually Impaired in a Controlled Area Using Deep Learning
Ahd Aljarf, Ghaidaa Almaghrabi, Halah Albarakati, Haneen Ahmed, Raghad Alharbi, Sawsan Aljuhani · 2022
With more than 285 million visually impaired people globally, there is a critical need for an assistive system that helps blind and visually impaired people to navigate easily. The EBSAR system is proposed to detect the objects using mobile video cameras and make a sound alert with the distance and right direction in Arabic language to the user. In the EBSAR system, the updated Convolutional Neural Networks (CNN) and You Only Look Once version four (YOLOv4) algorithms are used for object detection. Besides, the key matching as these algorithms give more accuracy. The camera of the phone is enough for detecting the objects and no special hardware is required. Thus, the system requires minimal effort from the user to use the system during everyday life.