Simple Smartphone-Based Guiding System for Visually Impaired People

Area of Science:

• Computer Science
• Assistive Technology
• Human-Computer Interaction

Background:

• Visually impaired individuals face significant navigation challenges, increasing collision risks.
• Existing navigation aids often lack real-time obstacle detection capabilities.
• Guidance is crucial for visually impaired people, especially in unfamiliar surroundings.

Purpose of the Study:

• To develop a smartphone-based guiding system for visually impaired individuals.
• To enhance environmental awareness and enable obstacle avoidance for improved navigation.
• To provide a user-friendly solution for independent travel for the visually impaired.

Main Methods:

• Integration of a computer image recognition system with a smartphone application.
• Implementation of two operating modes: online and offline, based on network availability.
• Utilizing advanced algorithms like Faster Region Convolutional Neural Network (Faster R-CNN) or You Only Look Once (YOLO) for obstacle recognition via backend server processing.

Main Results:

• The system successfully recognizes multiple obstacles in real-time from smartphone-captured images.
• Obstacle recognition accuracy reached 60%, deemed sufficient for practical assistance.
• The system provides visually impaired users with information on obstacle types and locations.

Conclusions:

• The proposed smartphone-based system offers a viable solution for navigation and obstacle avoidance for the visually impaired.
• The system enhances user awareness of surroundings, promoting safer and smoother travel.
• Further development could improve recognition accuracy and expand system functionalities for broader application.