AI-Enhanced Visual Guiding System Using Chat GPT for the Visually Impaired
Hossam A. Elshiekh
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Sara A. Ghaly *
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Khaled W. Khalaf
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Abdelrahman M. Elsayed
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Aya K. Fouda
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Kareem E. AboElhassan
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Sherif M. Maher
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Hend K. Rakha
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Sana A. Eissa
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Maha S. Bakr
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Abdallah E. Younes
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
Amira. A.Elsonbaty
Communication & Electronics Department, Higher Institute of Engineering and Technology, New Damietta, 34517, Egypt.
*Author to whom correspondence should be addressed.
Abstract
Introduction: Individuals with visual impairments encounter numerous challenges in their daily lives, as modern assistive devices frequently fail to meet consumer needs regarding affordability and functionality. This paper introduces a novel design for assistive smart glasses aimed at visually impaired people. The goal is to support various daily activities through a wearable format. As a proof of concept, this study focuses on application that text recognition technology that aids in reading printed materials. The overall cost is minimized by employing a Raspberry Pi 2 single-board computer for processing and a Raspberry Pi 2 camera for image capture.
Aims: To develop an AI-powered wearable visual guidance system for visually impaired individuals that provides real-time audio feedback based on environmental understanding. To integrate Chat GPT and computer vision for enhanced contextual awareness and conversational interaction. To assess the effectiveness, usability, and real-time performance of the system in assisting navigation and object recognition.
Type: Experimental prototype with field-testing.
Participants: 10–15 visually impaired volunteers.
Duration: 4–6 weeks for testing and data collection.
Environment: Both controlled indoor environments and real-world outdoor settings (e.g., sidewalks, shops).
Evaluation: Quantitative performance metrics and qualitative user feedback.
Methodology or Methods: The AI-Enhanced Visual Guidance System is a wearable solution designed to aid visually impaired individuals by providing real-time audio feedback based on visual input. It integrates camera-based perception, AI-powered scene understanding, and conversational interaction through Chat GPT. HARDWARE COMPONENTS :Raspberry Pi Zero 2 W (central processing unit) ,Pi Camera Module (image capture) ,Mini speaker or earphones (audio output), Portable power supply (battery pack) ,Micro SD card with pre-installed software stack .
Results: Results demonstrate that the system offers promising improvements in usability and real-time assistance, highlighting its potential to enhance mobility and independence for the visually impaired.
Conclusion: The proposed system revolves around the concept of smart assistive glasses for visually impaired individuals, utilizing artificial intelligence.
The implemented OCR system enables blind individuals to read text from boards by converting it into speech. As a result, the proposed solution serves as a powerful assistive tool for enhancing mobility and independence.
The system proved to be highly effective in enhancing the autonomy of visually impaired individuals. The integration of Chat GPT added a human-like, flexible layer of interaction. Despite minor technical limitations, the initial results are very promising and indicate strong potential for further development and real-world application.
Keywords: Visually impaired, text recognition, OCR, Raspberry pi 2, smart glass