Implementation of Color-Blind Aid System

Abstract

Color-blind is a physical defect causing patient loses the ability to recognize colors either particular or the whole of them. This disability is problematic in daily life, moreover in some specific areas that require carefulness eyesight. We propose a vision aid kit with improved user experience, such as finger pointer and voice command-and-response. Our work proofs the design and implementation of color-blind aid system for embedded and mobile device. This study used Windows Embedded Standard 2009, Windows Phone 7, Speech API library, Open CV library and EmguCV wrapper. The performance of each functionality is evaluated. From various testing that are conducted, the system can best detect color samples with precision level of 90.67% on embedded device implementation and 95.33% on mobile device implementation. The best fingertip detection rate attained is 89.6% with normal lighting condition. The voice command works with detection rate of 75.87%, meanwhile for the synthesized speech response, 88, 33% respondents can understood the words well.