LidSonic V2.0: A LiDAR and Deep-Learning-Based Green Assistive Edge Device to Enhance Mobility for the Visually...

Area of Science:

• Assistive Technology
• Computer Vision
• Robotics

Background:

• Over a billion people globally live with disabilities, with 253 million experiencing visual impairment or blindness.
• Existing assistive devices for visual impairment often lack maturity and fail to meet user needs.
• There is a critical need for affordable, low-power assistive technologies to enhance independence for visually impaired individuals.

Purpose of the Study:

• To propose and develop a novel, low-cost smart glasses system (LidSonic V2.0) for environment perception and navigation assistance for the visually impaired.
• To leverage LiDAR, ultrasonic sensors, and deep learning for efficient obstacle detection and object classification.
• To create an accessible and user-friendly assistive device with both haptic and auditory feedback.

Main Methods:

• The LidSonic V2.0 system integrates LiDAR, a servo motor, and an ultrasonic sensor with an Arduino Uno microcontroller on smart glasses.
• Data processing for obstacle detection is performed on the edge device, with a companion smartphone app providing object classification and spoken feedback.
• The system utilizes WEKA and TensorFlow platforms, employing affordable, off-the-shelf components costing under $80.

Main Results:

• LidSonic V2.0 demonstrates significantly reduced processing time and energy consumption compared to image-processing-based systems for obstacle classification.
• Prototype testing in real-world environments confirmed the system's effectiveness in environment perception and navigation assistance.
• The system provides efficient data processing, faster inference, and decision-making with lower energy usage and smaller data sizes.

Conclusions:

• The developed LidSonic V2.0 system offers an inexpensive, miniature, and effective solution for aiding the visually impaired.
• The proposed approach facilitates faster communication and decision-making, suitable for edge, fog, and cloud computing applications.
• This research encourages the development and adoption of affordable assistive technologies, improving the quality of life for individuals with visual impairments.