Text Signage Recognition (TSR) Noise Filtering Technique for Visually Impaired People (VIP)

Finding one’s way across campus to a class, locating room or offices in a large
medical centre, or just walking alone in the street to meet people has always been
difficult or impossible for the visually impaired (VIP). Most of the time they have to
be accompanied by someone or using a white cane and often have to use the wellknown
routes that they are familiar with. Independent navigation might be an easy
task for the normal sighted people but not VIP. Implementation of OCR on mobile
application is a great assistive technology that can assist the VIP during their
independent navigation. Referring to the previous TSR prototype application
developed in Android mobile application, we found that the usability of the prototype
application has encountered problems with the presence of noise from the image
captured by the VIP. Therefore, the author has proposed a new TSR technique to
filter the noise present in the image captured by VIP using Gaussian filter. The
project aims to deliver the proposed TSR technique to the targeted user which is low
visually impaired people after considering the user’s experience upon their
interaction with the existing TSR prototype application.