Eyes for the blind: Converting real-time images to speech

About 12 million people in the United States alone have a vision impairment, including 1 million who are blind. A visually impaired person relies on their other senses to understand their environments such as touch or hearing. It is almost impossible for a blind person to know the object that is in front of him without touching it. Sometimes physical contact can be dangerous to the person. This project tries to transform the visual word into an audio world that can be informative to a person who is visually impaired or blind.

There are existing systems in the market for helping the blind. However, in this project, we try to integrate object detection to navigate the blind person on the streets and help to identify general household objects including the capability to read out labels. This could especially be useful when trying to learn expiration dates or with medication.