Design and Empirical Investigation of Tactile Images for Accessing Photographs

Abstract

This Ph.D. thesis creates and empirically evaluates a system for making photographs accessible through touch, designed for situations where visual perception is limited. Given the reduced bandwidth of touch compared to vision, tactile representations are required to convey semantic information to allow meaningful interpretation by users. To address this, we developed Pic2Tac, an automated technical system that translates semantic information in photographs into a simplified binary output of symbols, or ``tactile words", using a small set of icons and patterns. Swell paper was selected as the tactile display medium for its accessibility, cost-effectiveness, and adaptability to ensure Pic2Tac remains practical and scalable for diverse communities facing a variety of unique challenges.

The thesis encompasses three projects:
* Initial development of Pic2Tac to detect main objects in photographs using computer vision and render them as simple tactile icons.
* Advancement of Pic2Tac to include background elements represented as tactile patterns, addressing some ambiguities in user narratives.
* A longitudinal study to assess whether repeated exposure to tactile images improves user comprehension and symbol recognition.

The results demonstrate that Pic2Tac effectively conveys semantic information through tactile words, with participants exhibiting increased comprehension and enjoyment over time. This research establishes foundational methods for improving tactile accessibility to photographs, paving the way for continued advancements in accessible media design.

Date of Award	10 Dec 2025
Original language	English
Awarding Institution	University of Bath
Supervisor	Peter Hall (Supervisor), Adwait Sharma (Supervisor) & Michael Proulx (Supervisor)