ISSN : 2663-2187

VIRTUAL SMART ASSISTANT FOR VISUALLY CHALLENGED PEOPLE FOR EMPOWERING INDEPENDENCE

Main Article Content

Dr T Sujatha, Dr Wilfred Blessing N R, Dr. SK WasimHaidar, Dr S Renuga S, Dr Suresh Palarimath
» doi: 10.48047/AFJBS.6.12.2024.101-111

Abstract

The Vision-Med: Virtual Assistant for Visually challenged is an application that assists visually impaired individuals in identifying their medications and ensuring their safety by preventing the consumption of expired drugs. The threat of expired drugs is a significant concern for visually impaired individuals who have difficulty reading the expiry date and name of the medicine on pharmaceutical packaging. Visually challenged individuals may require the assistance of a third person to help them with their medication every time. Expired drugs can lose their potency and effectiveness, leading to adverse reactions and health complications. For visually impaired individuals, the inability to read the expiration date of the medication can lead to confusion and medication errors. This occurs when a person accidentally consumes an expired drug, potentially leading to adverse health effects. The cost of replacing expired medication can be substantial, placing additional financial hardship on patients. The Vision Med application provides audio feedback to the user about the name and expiration date of the medication which leads to reducing the use of tablets that have expired. The application leverages Computer Vision (CV), Natural Language Processing (NLP) techniques, and Text to Speech (TTS) conversion to recognize medicine names and expiration dates from images of medication packaging. It provides audio feedback to the user about the medicine’s name and expiration date, improving medication adherence and enhancing the independence of visually impaired individuals. The Vision-Med application is a game changer for visually impaired individuals who struggle with medication management, thereby promoting safe medication and independent living. The performance of the application is evaluated with an accuracy of 98%.

Article Details