Euro Banknote Recognition System for Blind People

Area of Science:

• Computer Vision
• Assistive Technology
• Biomedical Engineering

Background:

• Visually impaired individuals face challenges in independently handling cash, particularly in distinguishing different denominations of Euro banknotes.
• Existing solutions for banknote recognition often lack portability or are not specifically designed for the needs of blind and visually impaired users.

Purpose of the Study:

• To develop and evaluate a wearable, portable system for the automatic detection and recognition of Euro banknotes for blind and visually impaired individuals.
• To enhance the financial independence and daily living capabilities of visually impaired people through accessible technology.

Main Methods:

• A portable system integrating a Raspberry Pi, a Raspberry Pi camera with No Infrared (NoIR) filter, and infrared illumination was developed.
• The system utilizes modified Viola and Jones algorithms for banknote detection and Speed Up Robust Features (SURF) technique for banknote value recognition.

Main Results:

• The system demonstrated a high accuracy rate of 97.5% for banknote value recognition.
• Banknote detection accuracy was recorded at 84%, indicating robust performance in identifying Euro banknotes.
• The device, embedded in sunglasses, allows for discreet and independent use by visually impaired individuals.

Conclusions:

• The developed portable system effectively enables blind and visually impaired users to detect and recognize Euro banknotes.
• This assistive technology significantly improves the ability of visually impaired individuals to manage cash transactions independently.
• The high accuracy achieved in recognition highlights the potential of this system for practical application in everyday scenarios.