A context-aware navigational autonomy aid for the blind

Alghamdi, S 2014, A context-aware navigational autonomy aid for the blind, Doctor of Philosophy (PhD), Computer Science and Information Technology, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
Alghamdi.pdf Thesis application/pdf 4.74MB
Title A context-aware navigational autonomy aid for the blind
Author(s) Alghamdi, S
Year 2014
Abstract Reaching a specific destination in unfamiliar environments is a real challenge for people with a vision disability, even if they use a white cane or guide dog. The focus of this thesis is to develop an integrated Navigational Autonomy Aid for the Blind (NAAB) system. The main target of NAAB is to produce a system that helps blind and visually impaired (BVI) people to navigate independently, particularly in unfamiliar indoor environments. The services that NAAB produce for BVI people include assistance to identify their current position precisely, guidance to their desired destinations and assistance with avoiding people who are likely to intersect their paths.

The major limitation of most navigation applications designed to assist BVI people in indoor environments is that each provides one specific service, such as positioning, navigation or obstacle avoidance. However, these applications rarely all incorporate these services, nor work effectively in indoor environments. Further, there are no commercially available integrated navigation systems to assist BVI people to navigate independently and reach their destination precisely in buildings they are visiting for the first time. This highlights the need to further address the research challenges.

Considering this problem, the main focus of this research is to investigate and develop an integrated navigation system that should be suitable to assist BVI people inside buildings and provide four services for them: positioning, navigation, avoidance of humans in their path (for blind users) and assistance with safe road crossing.

NAAB has three components: active radio frequency identification (RFID), quick-response (QR)-code and a 3D camera (Kinect). The integrated term refers to multiple technologies and multiple services. The first is positioning, which aims to assist BVI people to indicate their current location precisely. Positioning in NAAB systems could be done by RFID alone, with 2-m accuracy, but once the QR-code was integrated into the system directionality concerns with RFID were resolved and accuracy was increased as well. The developed RFID positioning algorithm is not based on received signal strength (RSS) but on a novel technique of positioning that has been developed as a combination of RSS with attenuation control level by the RFID reader, which is carried by the user.

The second service is navigation, which aims to assist BVI people to reach their desired destination. Navigation includes positioning based on a developed positioning algorithm, determining the most preferable path for the user from their current location to the destination through the shortest path in most cases, and detection of deviation of the user from the suggested path to recalculate the path to the destination from the wrong location, as far as is possible. The efficiency of the proposed navigation system (NAAB) was tested on blind participants and their perspectives regarding the system were taken into account to improve the usability of the system.

The third service uses the Kinect, where NAAB can detect moving people in the path of the user and predict the movement of each person to warn the user about any potential intersection with these people. service.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Computer Science and Information Technology
Keyword(s) Blind Assistance
Navigation and positioning
RFID
QR-code
Kinect
Versions
Version Filter Type
Access Statistics: 387 Abstract Views, 794 File Downloads  -  Detailed Statistics
Created: Fri, 14 Nov 2014, 13:47:43 EST by Maria Lombardo
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us