VisionPal: Visual Assistant System for the Visually Impaired People
Visual impairment affects millions of individuals worldwide, hindering their ability to interact with society. This research introduces VisionPal, a comprehensive smartphone-based application for the Visually Impaired. The system employs advanced image processing and deep learning techniques to addr...
Saved in:
| Published in: | 2023 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems (ICSES) pp. 1 - 8 |
|---|---|
| Main Authors: | , , , , , |
| Format: | Conference Proceeding |
| Language: | English |
| Published: |
IEEE
14-12-2023
|
| Subjects: | |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Visual impairment affects millions of individuals worldwide, hindering their ability to interact with society. This research introduces VisionPal, a comprehensive smartphone-based application for the Visually Impaired. The system employs advanced image processing and deep learning techniques to address the challenges visually impaired individuals face in outdoor environments. The core of VisionPal is a smartphone application that integrates features such as obstacle detection, vacant seat identification, road sign recognition, and pedestrian crossing assistance. The background and literature survey delve into related works that are focused on this study. Leveraging YOLOv8, InceptionV3, Tesseract, and MobileNetV2 models, the methodology incorporates machine learning models to detect obstacles, vacant seats, crowd levels, road signs, and color lights. Researchers developed a mobile application allowing users to interact via vocal commands. Testing reveals high accuracy, and the findings depict that the YOLOv8 model precisely identifies vacant seats and crowd levels, achieving an overall accuracy of 98.74%, and obstacle identification depicts an overall accuracy of 94.38%. Further, the InceptionV3 model exhibits 97% accuracy in recognizing road signs and colors. The MobileNetV2 model achieves 98.5% accuracy in identifying traffic lights and pedestrian crossings. User feedback emphasizes the user-friendliness and effectiveness of the application in enhancing outdoor experiences for visually impaired individuals. The significance of this work is its ability to empower visually impaired individuals, enhancing their autonomy, safety, and quality of life. VisionPal addresses real-world challenges by offering a reliable and innovative solution for the visually impaired to navigate public places. The integration of state-of-the-art technologies makes VisionPal a promising tool for promoting inclusion and enabling visually impaired individuals to engage more fully in society. |
|---|---|
| DOI: | 10.1109/ICSES60034.2023.10465537 |