Portable Camera Based Text Reading of Objects for Blind Persons

摘要

This paper proposes a portable camera-based assistive text reading framework to help blind persons to read text labels and product packaging from hand-held objects in their daily lives. The system framework consists of three functional components: First, scene capture-using a mini camera, the text which the user needs to read gets captured as an image and has to be sent to the image or data processing platform., second, data processing - where text will be filtered from the surrounding and will be recognized by optical character recognition (OCR) software, and finally, Speech output - A filtered text will be passed into this system to get an audio output. In addition to text recognition using OCR, the project also includes template matching technique as a separate method for recognizing certain objects like currency notes. Since the project should be portable to assist the blind persons, the entire application is based on Raspberry Pi.