Related Experiment Video
Updated: Apr 21, 2026
05:36
Subjective Refraction Test Using a Smartphone for Vision Screening
Published on: October 18, 2024
2.1K
SMARTPHONE-BASED CROSSWALK DETECTION AND LOCALIZATION FOR VISUALLY IMPAIRED PEDESTRIANS.
Vidya N Murali1, James M Coughlan1
1The Smith-Kettlewell Eye Research Institute, San Francisco, CA.
... IEEE International Conference on Multimedia and Expo Workshops. IEEE International Conference on Multimedia and Expo
|November 1, 2014
Summary
This study introduces Crosswatch, a smartphone system aiding visually impaired travelers at intersections. It enables precise self-localization using 360° panoramas and aerial map matching for safe navigation.
Area of Science:
- Computer Vision
- Human-Computer Interaction
- Assistive Technology
Background:
- Existing computer vision methods for blind travelers struggle with framing crosswalks due to visual limitations.
- Accurate self-localization at traffic intersections is crucial for safe navigation for visually impaired individuals.
Purpose of the Study:
- To develop and validate a novel computer vision-based system for self-localization and crosswalk detection for blind and visually impaired travelers.
- To overcome the limitations of single-frame imaging in assistive navigation systems.
Main Methods:
- Utilizes a smartphone system (Crosswatch) to capture 360° image panoramas.
- Converts panoramas to an aerial (overhead) view, centered on the user's position.
- Matches the aerial view with satellite imagery templates (e.g., Google Maps) for crosswalk detection and self-localization.
Main Results:
- Demonstrates the feasibility of precise user self-localization relative to crosswalks using the developed method.
- Successfully detects crosswalk features and estimates user position from imagery acquired by blind users.
Conclusions:
- The Crosswatch system shows promise in enhancing safe navigation for visually impaired individuals at traffic intersections.
- The approach of converting 360° panoramas to aerial views for map matching is effective for assistive navigation.
