Improving exploration of tactile graphics by visually impaired people: Theoretical advances and a novel mobile application

Zeinullin, Maralbek (2025) Improving exploration of tactile graphics by visually impaired people: Theoretical advances and a novel mobile application. PhD thesis, University of Glasgow.

Abstract

The goal of this PhD study is to improve the accessibility of tactile graphics for people with visual impairments. Tactile graphics provide a means for blind individuals to understand nontextual information through touch. However, they often require supplementary audio or Braille text descriptions, which can be time-consuming and create obstacles to learning. For instance, some individuals may find that reading Braille or listening to a screen reader takes longer because they have to go through all the text or audio to locate the necessary information. To address these challenges, this study develops and evaluates TAURIS (Tactile AUdio Responsive Intelligent System), a novel system that provides real-time audio descriptions that accompany tactile graphics. TAURIS is comprised of pre-labelled tactile graphics, an interactive web tool for labelling, and a mobile application. The mobile application relies on a customised deep learning computer vision model to detect the user’s fingertips as they explore the tactile graphics and provides information about what they are touching. Notably, no system simultaneously detecting the fingertips of both hands in real-time and under different lighting conditions has been implemented on a mobile device.

In order to evaluate the efficacy of TAURIS, a mixed-methods approach was employed. This approach consisted of structured interviews, experiments, and post-experimental feedback sessions involving a total of 20 participants, including one pilot participant. The structured interviews were designed to gather information on the participants’ experiences with tactile graphics and mobile devices. In the subsequent experiments, response times and the number of correct answers were recorded during end-user testing of the app. The resulting data was analysed using the Wilcoxon signed-rank test. The analysis revealed three statistically significant findings: TAURIS allowed for faster interaction with tactile graphics and higher accuracy in answering questions compared to both Braille text and screen readers. Additionally, while no significant difference in memory retention was observed between TAURIS and Braille text, TAURIS demonstrated a significant advantage in memory retention over screen readers. Additionally, the participants provided feedback on their impressions, comments, and suggestions for improving the system. Finally, a mixed-methods approach was used to triangulate the data from multiple sources and strengthen the validity of these findings.

Based on these findings, TAURIS has the potential to empower individuals with visual impairments by providing an accessible and efficient tool that supports independent learning and improves knowledge retention. Further research with a larger number of participants in various schools for the blind and countries would be valuable for gaining additional insights, increasing the power of the statistical analysis, and enabling further comparisons.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Additional Information: Supported by funding from the Bolashak scholarship program.
• Subjects: T Technology > T Technology (General)
• Colleges/Schools: College of Science and Engineering > School of Engineering > Biomedical Engineering
• Supervisor's Name: Hersh, Dr. Marion
• Date of Award: 2025
• Depositing User: Theses Team
• Unique ID: glathesis:2025-84854
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 30 Jan 2025 14:36
• Last Modified: 30 Jan 2025 14:50
• Thesis DOI: 10.5525/gla.thesis.84854
• URI: https://theses.gla.ac.uk/id/eprint/84854

Related URLs

• Article DOI