An improved object identification for NAVI

Abstract

Navigation assistance for visually impaired (NAVI) is a vision substitute system designed to assist blind people for autonomous navigation. NAVI working concept is based on 'image to sound' conversion whereby input image captured from the video sensor is processed and transformed into stereo sound pattern. The sound is conveyed to blind's auditory system via stereo headphones. Processing in NAVI can be divided into two sub modules, which are the vision module and sonification module. In the vision module, image processing is employed to identify objects from the image. This paper presents an improved object identification methodology for NAVI. Objects are identified by its closed boundary. Properties of each object are evaluated using AI technique as to assign vision preference, so that blind user can determine the location as well as other properties of that object. Experimentation and results are presented to evaluate the proposed methodology for blind navigation purposes.