The goal proposed project is to develop an indoor navigation system for persons with visual impairment using computer vision and machine learning. Here, an intelligent decision support system (DSS) based on artificial intelligence i.e. machine learning will be developed to reliably sense the environment by using cameras, translate the information, navigate and suggest personalized decisions to persons with visual impairments. Main contributions of this projects are in twofold: 1) Vision-based Feature Extraction: RGB camera with RFID tags i.e. labelled objects, common signs, product logos will be used for indoor navigation. 2) Decision Support System: where features from previous task will provide as input to a hybrid case-based reasoning (CBR) approach together with other methods e.g., deep learning, and fuzzy logic to provide an experience based personalized decision support.
A Vision based Indoor Navigation System for Individual with Visual Impairment (May 2020) Mobyen Uddin Ahmed, Mohammed Ghaith Altarabichi, Shahina Begum, Fredrik Ginsberg , Robert Glaes , Magnus Östgren , Hamidur Rahman, Magnus Sörensen International Journal of Artificial Intelligence (IJAI)