Integrated assistance system for blind

Abstract

This paper proposes an integrated assistance system for blind and visually impaired people. The system incorporates three basic features viz, color detection, text detection and location finding. The process of color detection includes capturing the objects through a webcam and applying the Color Image Processing techniques. To assist the person in reading, text to speech conversion method is used. Initially, the printed document is captured through webcam, then adaptive Gaussian thresholding, dilation and erosion techniques are applied to detect the text. Finally the detected text is mapped to the corresponding predefined words, which are then converted to speech. For extracting text Tesseract OCR Engine is used. In addition a GPS module is added for tracking the location of the blind person and to send the location details in the form of an SMS to the registered phone numbers. The entire system is implemented using Raspberry Pi.