Object Recognition System for the Visually Impaired: A Deep Learning Approach using Arabic Annotation

Object detection is an important computer vision technique that has increasingly attracted the attention of researchers in recent years. The literature to date field introduced a range object models. However, these models have largely been English-language-based, and there only limited number published studies addressed how can be implemented for Arabic language. As far as we are aware, generation text-to-speech engine utter objects’ names their positions images help Arabic-speaking visually impaired people not investigated previously. Therefore, this study, propose segmentation model based on Mask R-CNN algorithm capable identifying locating different objects images, then uttering Arabic. proposed was trained Pascal VOC 2007 2012 datasets evaluated testing set. We believe one few uses train test model. performance compared with previous literature, results demonstrated its superiority ability achieve accuracy 83.9%. Moreover, experiments were conducted evaluate incorporated translator TTS engines, showed could effective helping understand content digital images.