A Systematic Review of Urban Navigation Systems for Visually Impaired People

Area of Science:

• Assistive Technology
• Computer Vision
• Machine Learning

Background:

• Blind and visually impaired people (BVIP) encounter significant challenges during outdoor navigation.
• Recent advancements in computer vision and machine learning offer potential solutions for enhanced pedestrian assistance.

Purpose of the Study:

• To conduct a comprehensive systematic review of research relevant to assistive outdoor navigation for BVIP.
• To analyze existing navigation methods, datasets, and identify current limitations across different navigation phases and tasks.

Main Methods:

• Systematic review of research articles focused on assistive outdoor navigation for BVIP.
• Categorization of navigation into distinct phases and tasks for structured analysis.
• Evaluation of research based on methods, datasets, and identified limitations.

Main Results:

• Identification of various assistive devices and technological approaches, including deep learning.
• Analysis of the state-of-the-art in environment interpretation and navigation assistance for BVIP.
• Overview of both commercial and non-commercial navigation applications available for BVIP.

Conclusions:

• The review provides a structured analysis of the current research landscape in assistive outdoor navigation for BVIP.
• It highlights the state-of-the-art and offers guidance for future research directions.
• This work aims to inform researchers and stakeholders for informed decision-making in the field.