LineChaser: A Smartphone-Based Navigation System for Blind People to Stand in Line

Title of the Paper

LineChaser: A Smartphone-Based Navigation System for Blind People to Stand in Lines

Paper Information

• Research Area: Accessibility Technology and Intelligent Navigation Systems
• Keywords: Visual impairment, positioning and navigation, pedestrian detection, queue navigation, human-computer interaction

Research Background and Problem

• Issues and Challenges:

  • Blind individuals face difficulties in locating the end of a queue and maintaining proper following of the person ahead in public spaces.
  • Social distancing requirements during the COVID-19 pandemic have further exacerbated the challenges for blind people to perceive queue dynamics.
  • Existing navigation technologies primarily address static target positioning, while navigation for dynamic queue ends has not been adequately explored.

• Significance:

  • Queuing is a common social behavior in daily life, and solving queuing issues for visually impaired individuals can significantly improve their social participation and independence.

• Research Motivation and Related Work:

  • Current navigation systems like Google Maps and BlindSquare rely on static route maps and are ineffective in addressing dynamic queue-end navigation.
  • Some computer vision-based systems can detect pedestrians but are not optimized for queuing scenarios.
  • Existing research includes developments in robotic queuing and obstacle avoidance technologies, but there is still a lack of queue navigation systems specifically designed for blind individuals.

Solution

• Research Method and System Design:

  • The authors developed a smartphone application—LineChaser—that utilizes the smartphone's RGB camera and infrared depth sensor to detect and track people in a queue.
  • The system employs audio and vibration feedback to guide users in correctly joining the queue and maintaining social distance.

• Innovations:

  • Combines dynamic path navigation with real-time pedestrian detection technology, leveraging smartphones as the hardware platform to enhance applicability.
  • Introduces a new target tracking algorithm that uses color histograms to distinguish target individuals in the queue, reducing the likelihood of tracking errors.

• Implementation Steps:

  1. Map Preparation and Positioning:
     • Use ARKit and pre-set AR markers to create a queue information map.
     • Users scan AR markers to locate their current position.
  2. Queue-End Detection:
     • The system detects nearby individuals, determines whether they are queuing, and identifies the last person in the queue.
     • Acquires the target individual's color histogram for subsequent tracking.
  3. Queue Following:
     • Follow the target at a specified social distance (1.7 meters) and guide the user to move within the queue.
  4. Audio and Vibration Feedback Integration:
     • Audio feedback provides navigation instructions (e.g., "Move towards 2 o'clock direction, 2.1 meters away").
     • Vibration feedback prompts the user to move or stop, offering additional information for understanding.

Research Outcomes

• Specific Results:

  • Two experiments were conducted with 6 and 12 completely blind participants, respectively.
  • LineChaser successfully assisted all participants in locating the queue-end and completing the queue-following task while maintaining appropriate social distance.

• Comparison with Existing Technologies:

  • The prototype system had a higher rate of tracking errors, whereas LineChaser resolved this issue using color histograms.
  • Compared to traditional navigation technologies, LineChaser provides the capability to locate dynamic queue ends.

• Experimental and Evaluation Results:

  • Participants correctly stayed in acceptable positions within the queue in 91.7% of cases.
  • The system's usability was well-received (SUS average score of 83.9, rated as "A"), significantly boosting users' confidence and comfort in queuing.

• Limitations and Future Directions:

  • The current system requires users to hold a smartphone, which may lead to fatigue; future iterations could explore wearable devices as alternatives.
  • Limitations of AR positioning technology may result in positioning errors, requiring further optimization or integration with other technologies.
  • Social acceptance needs to be assessed, especially as camera-based systems may raise privacy concerns.
  • Plans to integrate with other navigation systems and conduct tests in real-world scenarios to further validate system performance and applicability.

Through the above analysis, LineChaser not only provides a practical solution to assist blind individuals in completing queuing tasks but also demonstrates the potential of combining low-cost and high-usability technology using smartphones.