Crosswalk Guidance System for the Blind

Area of Science:

• Assistive technology
• Human-computer interaction
• Robotics

Background:

• Street crossing presents significant mobility challenges for visually impaired individuals, particularly in urban settings.
• Existing solutions primarily focus on detecting crosswalks and signals, with limited consideration for pedestrian mobility during crossing.
• There is a need for advanced systems to support safe and independent street crossing for the visually impaired.

Purpose of the Study:

• To develop and evaluate a wearable goggle system for guiding visually impaired pedestrians across streets.
• To integrate real-time detection of crosswalk signals, path planning, and verbal guidance functionalities.
• To assess the system's effectiveness and reliability in real-world outdoor crosswalk scenarios.

Main Methods:

• A commercially available wearable goggle system was programmed with algorithms for crosswalk signal detection and path planning.
• The system was designed to provide real-time verbal guiding cues with semantic features for user navigation.
• Verification testing included evaluating signal detection accuracy and conducting trials with visually impaired subjects.

Main Results:

• Crosswalk signal detection demonstrated reliability, though hyper-parameter tuning was sometimes necessary to minimize false positives.
• The wearable system successfully guided visually impaired subjects across an outdoor crosswalk.
• Testing confirmed the potential of the system to aid in independent street crossing.

Conclusions:

• The developed wearable goggle system offers a promising solution for enhancing the mobility and safety of visually impaired pedestrians.
• Real-time guidance and path planning integrated into a wearable system can effectively support visually impaired individuals during street crossing.
• Further refinement of detection algorithms may improve system performance in complex environments.