Chart4Blind: An Intelligent Interface for Chart Accessibility Conversion

TL;DR

Chart4Blind presents an interactive web-based interface that semi-automates the conversion of bitmap line charts into accessible formats (SVG, CSV, and descriptive text) through an Data Extraction Module (line segmentation via instance segmentation and OCR) and a Rendering Module (real-time SVG with accessibility-aware outputs). The system targets tactile, screen-reader, and embossed printing modalities, and is guided by need-finding interviews and usability studies that report an average SUS of 90% and a ~4-minute completion time for inexperienced users. Key contributions include an end-to-end pipeline for metadata extraction, semi-automatic data calibration, and template-based chart summarization, enabling broader inclusion of BVI users in data visualization workflows. The work demonstrates practical impact by reducing manual effort, enabling collaborative annotation, and providing open-source access to tools and demos, with future work extending support to additional chart types and richer natural language descriptions.

Abstract

In a world driven by data visualization, ensuring the inclusive accessibility of charts for Blind and Visually Impaired (BVI) individuals remains a significant challenge. Charts are usually presented as raster graphics without textual and visual metadata needed for an equivalent exploration experience for BVI people. Additionally, converting these charts into accessible formats requires considerable effort from sighted individuals. Digitizing charts with metadata extraction is just one aspect of the issue; transforming it into accessible modalities, such as tactile graphics, presents another difficulty. To address these disparities, we propose Chart4Blind, an intelligent user interface that converts bitmap image representations of line charts into universally accessible formats. Chart4Blind achieves this transformation by generating Scalable Vector Graphics (SVG), Comma-Separated Values (CSV), and alternative text exports, all comply with established accessibility standards. Through interviews and a formal user study, we demonstrate that even inexperienced sighted users can make charts accessible in an average of 4 minutes using Chart4Blind, achieving a System Usability Scale rating of 90%. In comparison to existing approaches, Chart4Blind provides a comprehensive solution, generating end-to-end accessible SVGs suitable for assistive technologies such as embossed prints (papers and laser cut), 2D tactile displays, and screen readers. For additional information, including open-source codes and demos, please visit our project page https://moured.github.io/chart4blind/.

Figures (6)

• Figure 1: The pipeline of Chart4Blind consists of the input of a bitmap line chart, followed by the Data Extraction Module, which includes an AI-based line segmentation and optical character recognition step, and a manual correction step by a sighted user (a). The Rendering Module updates the information in real-time and ensures an accessible representation (c). The system allows the export of the information to an SVG and a CSV format. The SVG can be accessed with a screen reader or printed as a tactile graphic (d). The metadata can be exported as accessible CSV as well.
• Figure 2: Rendering Module view for a bitmap line chart. (a) displays the digitally accessible SVG view, ideal for screen readers, and refreshable tactile devices. (b) shows the print-accessible SVG view, suitable for print modalities such as embossed papers or laser cut.
• Figure 3: An overview of Chart4Blind interface sections: (a) Home menu for actions like upload, undo, redo, and tutorials. (b) AI toolbar with OCR and segmentation models. (c) Canvas for the uploaded chart, allowing interaction for calibration points and predicted line adjustments. (d) Rendering Module for real-time SVG visualization before export. (e) Metadata section for visualizing extracted line data and seamlessly drag-and-drop of textual content.
• Figure 4: Four line charts with different complexities utilized for the user study: Simple charts (a) and (b) each contain one simple line trend for the tutorial and main session respectively. (c) A compound chart with additional lines overlapped, and visible axes. (d) A dense chart featuring relatively complex trends, point annotations, and less visible axes.
• Figure 5: On the left, average conversion task completion time in minutes:seconds. On the right, a radar chart depicting the number of clicks for the top 5 sections interacted with in Chart4Blinds.