Enhanced navigation systems in GPS denied environments for visually impaired people: A Survey

Ali Hojjat

Abstract

Although outdoor navigation systems are mostly dependent on GPS, indoor systems have to rely upon different techniques for localizing the user, due to unavailability of GPS signals in indoor environments. Over the past decade various indoor navigation systems have been developed. In this paper an overview of some existing indoor navigation systems for visually impaired people are presented and they are compared from different perspectives. The evaluated techniques are ultrasonic systems, RFID-based solutions, computer vision aided navigation systems, and smartphone-based applications.

Introduction

Navigating unfamiliar indoor spaces presents significant challenges for the visually impaired (BVI), primarily due to a lack of "preview" capabilities, limited environmental layout knowledge, and a lack of orientation information. While traditional aids like long canes and guide dogs are essential, they cannot detect overhead obstacles or provide turn-by-turn directions. Because GPS signals cannot penetrate building walls, indoor environments remain "GPS-denied," necessitating alternative localization and guidance technologies. This survey categorizes these technologies into micro-navigation (immediate obstacle avoidance) and macro-navigation (localization and path planning).

Methodologies in Indoor Navigation

The survey reviews five distinct technological approaches to solving the indoor wayfinding problem:

1. Ultrasonic-Based Systems

These systems utilize wearable sensors and augmented canes to detect obstacles via 40Hz sound waves. A notable feature is the "record and playback" mode, allowing users to save and retrace specific routes.

2. RFID-Based Solutions (PERCEPT)

The PERCEPT system relies on passive RFID tags installed on infrastructure (doors, kiosks). Users interact with these tags via specialized gloves, which communicate with a smartphone via Bluetooth to provide audio feedback.

3. Infrared Triangulation

This approach uses an augmented white cane equipped with IR LEDs. Fixed "Wiimote" cameras mounted on walls track these LEDs. The data is processed by a central computer to triangulate the user's position within a specific range, eventually providing audio feedback to the user via a smartphone.

System workflow for infrared-based indoor navigation, showing the communication loop from the user to the smartphone.

4. Computer Vision (CV)

CV systems employ cameras and geometric algorithms to identify architectural features like doors. Optical Character Recognition (OCR) is used to read signage, enabling the system to distinguish between identical-looking entryways.

5. Smartphone and NFC (PERCEPT II)

An evolution of RFID systems, PERCEPT II replaces specialized hardware with Near Field Communication (NFC) tags and standard smartphones. This shift leverages built-in accessibility features and reduces the hardware burden on the user.

Technical Analysis of Localization

A key technical challenge is precise localization. The survey identifies four primary categories: dead reckoning, direct sensing, triangulation, and pattern recognition. Triangulation, as seen in IR systems, offers high precision but is limited by the "triangulable area" defined by the placement of fixed sensors.

The triangulable area and user position projection between two fixed Wiimote sensors.

Key Insights and Results

The survey highlights several critical findings regarding the current state of Electronic Travel Aids (ETAs):

  • Infrastructure Trade-offs: High-precision systems (IR, RFID) require "instrumenting" buildings, which poses scalability and cost challenges.
  • Interaction Design: Localization accuracy is secondary to the user interface. Effective systems must utilize haptic (vibration) or audio (speech synthesis) feedback to be truly accessible.
  • Performance Metrics: Ultrasonic systems offer a detection range of 0.03m to 6m with a 72-degree field of view. CV systems excel at semantic identification (reading signs) where sensor-only systems fail.

Conclusion and Future Work

The primary bottleneck for current indoor navigation systems is scalability. Future research is moving toward smartphone-centric solutions that utilize internal sensors (accelerometers, gyroscopes) to minimize external hardware requirements. Additionally, robust computer vision algorithms capable of handling real-world lighting and clutter without pre-installed markers remain a priority for the field.

Summary updated

Abstract

Although outdoor navigation systems are mostly dependent on GPS, indoor systems have to rely upon different techniques for localizing the user, due to unavailability of GPS signals in indoor environments. Over the past decade various indoor navigation systems have been developed. In this paper an overview of some existing indoor navigation systems for visually impaired people are presented and they are compared from different perspectives. The evaluated techniques are ultrasonic systems, RFID-based solutions, computer vision aided navigation systems, ans smartphone-based applications.
BibTeX
@misc{hojjat2018enhanced,
  author = {Ali Hojjat},
  title = {{Enhanced navigation systems in GPS denied environments for visually impaired people: A Survey}},
  year = {2018},
  eprint = {1803.05987},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1803.05987},
}
Figure