Haptic Gloves Prototype for Audio-Tactile Web Browsing
Andrii Soviak · 2015 · ASSETS '15: Proceedings of the 17th International ACM SIGACCESS Conference on Computers & Accessibility · doi:10.1145/2700648.2811329
Summary
This paper presents preliminary work on FeelX, a haptic gloves system designed to enable blind users to explore web page layouts through touch. The research addresses a fundamental limitation of screen readers: they present web content serially, creating a one-dimensional mental model that prevents blind users from understanding the two-dimensional spatial relationships between page elements. Understanding where elements are located and how they relate spatially helps sighted users navigate efficiently, but this information is largely inaccessible to screen reader users. The FeelX system consists of haptic gloves worn by the user, an overhead camera for finger tracking, and software that interfaces with web browsers and screen readers. Each finger (except the thumb) is equipped with a Braille cell that can render a 2x4 tactile grid with dots spaced 2.5mm apart, positioned on the pad of the fingertip (the most sensitive part of the finger). The system tracks finger positions using infrared LEDs placed on each fingertip and mid-finger, detected by a webcam converted to infrared by removing its IR filter, with OpenCV handling the computer vision processing. As the user moves their hands across a flat surface like a desk, the system maps finger positions to corresponding locations on the web page and renders appropriate tactile feedback. The gloves can display both Braille characters and simple shapes like lines and boxes—useful since most websites are styled with rectangular sections. Pressure sensors on each finger enable clicking or selection actions. The system works alongside existing screen readers, which can be triggered to narrate specific content when the user explores regions of interest.
Key findings
The paper presents a system architecture and prototype implementation rather than user study results. The key technical contributions include: The finger tracking approach using dual infrared LEDs per finger (tip and middle) enables tracking of both position and pointing direction, an improvement over simpler single-LED approaches that only recognize basic gestures. The use of standard Braille cells for tactile output leverages existing, well-understood hardware while the computer vision approach uses commodity components (webcam, OpenCV library). The architecture integrates several components: the Camera captures video of the gloves; the Finger Tracker resolves X,Y coordinates for each finger; the Interface Manager obtains page geometry via a browser plug-in and maps finger positions to screen coordinates; the Controller sends tactile rendering commands to the gloves; and pressure sensors enable input from the user back to the system. Compared to existing alternatives, the author argues FeelX offers several advantages. Pin-matrix tactile displays like BrailleDis9000 (120x60 resolution) are expensive, physically small, and have limited resolution. FeelX uses fewer Braille cells (only on fingertips), making it cheaper and more portable, while the "tactile area" is limited only by the desk surface and camera viewing angle rather than device size.
Relevance
This work addresses an important gap in web accessibility: spatial understanding. While screen readers provide access to web content, the loss of spatial context forces blind users into inefficient serial exploration. Research has shown that blind users spend excessive time on simple web tasks partly because they cannot quickly scan and orient themselves within page layouts the way sighted users do. The haptic gloves approach is novel in that it decouples tactile feedback from a fixed display surface, allowing users to explore larger virtual spaces using natural hand movements across any flat surface. This could potentially provide more intuitive spatial exploration than stationary tactile displays. As a preliminary/conceptual paper, significant work remains before practical deployment. No user evaluation with blind participants is reported, so usability, learnability, and actual benefits for web navigation are unvalidated. Open questions include: how well can users build accurate mental models of page layouts through this interface? What is the learning curve? How does cognitive load compare to screen-reader-only browsing? The prototype also requires an overhead camera setup that may limit portability. Nevertheless, the work provides a foundation for exploring haptic-augmented web accessibility and could inform future research on spatial understanding for blind users.
Tags: blindness · haptic interface · web accessibility · screen reader · tactile display · wearable · Braille · spatial layout · assistive technology