Assistive System For Product Label Detection With Voice Output For Blind Users

We propose a camera-based assistive text reading framework to help blind persons read text labels and product packaging from hand-held objects in their daily lives. We first propose an efficient and effective motion based method to define a region of interest (ROI) in the video by asking the user to shake the object. To automatically localize the text regions from the object ROI, we propose a novel text localization algorithm by learning gradient features of stroke orientations and distributions of edge pixels in an Ada boost model. Text characters in the localized text regions are then binarized and recognized by off-the-shelf optical character recognition (OCR) software. The recognized text codes are output to blind users in speech. Keywords-Assistive text, blind persons, distribution of edge pixels, hand-held objects, optical character recognition (OCR), stroke orientation, text localization.