Abstract
One of the most challenges that Visually Impaired patients face in their daily life activities is linked to drug identification, and there are not sufficient special pharmaceuticals services provided to help them in identifying the different medications they use as reported by a research conducted in Saudi Arabia [3]. This make them depend on the help of people with normal vision around them, or develop their own systems like identifying a medicine by its odor or putting medication boxes with similar shapes in different places. The difficulty that visually impaired patients find in identifying their medication boxes may lead them to take wrong medications sometimes. This may highly affect their health, especially if they are suffering from a chronic disease and need to follow their medications regularly. There are some solutions for helping such patients in identifying their medication boxes like barcode or Near Field Communication (NFC) readers. The existing solutions are either unaffordable, require carrying additional devices, or difficult to use. Developing and designing more usable and accessible software solutions for visually impaired users became possible with the technologies such as computer vision. In this paper, we have focused on leveraging of the possibility offered by this field to develop a system that enables the patients to identify their medication without the need of any equipment other than their mobile phones. After developing our proposed system, we have evaluated its usability and accessibility with visually impaired users, and then presented the results and discussed some limitations and potential future works.