Real-Time Walk Light Detection with a Mobile Phone

Area of Science:

• Assistive Technology
• Computer Vision
• Human-Computer Interaction

Background:

• Urban intersection crossings pose significant dangers for blind and visually impaired individuals.
• Existing assistive technologies often focus on navigation and alignment, with less attention to the critical timing of crossing.

Purpose of the Study:

• To develop and validate a portable system that alerts visually impaired users in real-time when a pedestrian Walk light is illuminated.
• To address the complementary challenge of knowing when to cross, building upon prior work on crosswalk alignment.

Main Methods:

• A prototype system was developed as a software application on a Nokia N95 mobile phone.
• Computer vision algorithms analyzed real-time video from the phone's camera to detect the Walk light.
• An audio tone alert was triggered upon successful detection of the Walk light.

Main Results:

• The system successfully detected the Walk light in real-time during experiments.
• A blind volunteer subject was effectively alerted when the Walk light appeared at urban intersections.
• Proof of concept was demonstrated for the portable Walk light detection system.

Conclusions:

• The developed portable system offers a promising solution for enhancing the safety of urban intersection crossings for visually impaired pedestrians.
• Real-time audio alerts based on visual Walk light detection can significantly improve situational awareness and crossing decisions.
• This technology represents a step forward in assistive technology for independent mobility.